Summary of the invention
Technical problem to be solved by this invention provides a kind of method and system of human body organ three-dimensional surface rebuilding, has the problem that simulation is rough, can not rebuild many bifurcateds complex organ cavity to solve existing housing surface method for reconstructing.
For solving the problems of the technologies described above,, the invention discloses following technical scheme according to specific embodiment provided by the invention:
A kind of method of human body organ three-dimensional surface rebuilding comprises:
Gather the spatial positional information of organ surface some spots;
Add one or more obligatory points that comprise spatial positional information;
According to the spatial positional information of described sampled point and obligatory point, construct and find the solution a function;
The point that all are identical with the functional value of sampled point carries out contour surface and extracts, and obtains the reconstruction curved surface of described organ.
Wherein, described obligatory point can perhaps spatial inner interpolation of sampled point institute envelope, perhaps add the spatial inside and outside of sampled point institute envelope simultaneously spatial outside interpolation of sampled point institute envelope.
Wherein, the spatial positional information of described obligatory point calculates acquisition according to the spatial positional information of sampled point, and perhaps the mode of selecting by man-machine interactive obtains.
Wherein, if ramose organ is arranged, then the locus of the obligatory point of described branch part can obtain by sampling.
Wherein, when constructed fuction, set the functional value of sampled point and obligatory point, wherein the functional value of sampled point is identical, and the functional value of obligatory point carries out gradual change according to its position in the spatial dimension of sampled point institute envelope.
Wherein, described function is formed by one group of RBF linear superposition.
Preferably, described contour surface is extracted by the lattice simplified method of triangular facet.
The lattice simplified method of described triangular facet comprises: set up a spatial three-dimensional grid of containing sampled point institute envelope; Calculate the functional value of described function on all grid vertexes; According to the difference of grid vertex functional value, determine the equivalent implicit surface identical with the sampled point functional value the cell of process; Determined all cells are calculated, solve described equivalent implicit surface and described cell and intersect the one or more triangular facets that form, the final contour surface that obtains by the triangular facet grid representation, i.e. organ curved surface of Chong Jianing.
Wherein, described sampled point can be in the inner surface collection of organ cavity, also can be in the outer surface collection of organ cavity.
Wherein, the collection of described sampled point can be passed through the detection method collection of electromagnetic field, microwave or sound wave.
Preferably, the quantity of described sampled point is 20~200.
Preferably, the quantity of described obligatory point is 6~70.
A kind of system of human body organ three-dimensional surface rebuilding comprises:
Information collecting device is used to gather the spatial positional information of organ surface some spots;
Information processor comprises:
Construction of function and find the solution the unit is used to add one or more obligatory points that comprise spatial positional information, and according to the spatial positional information of described sampled point and obligatory point, constructs and find the solution a function;
The curve reestablishing unit is used for the point that all are identical with the functional value of sampled point and carries out the contour surface extraction, obtains the reconstruction curved surface of described organ.
Wherein, described construction of function and find the solution spatial outside the add obligatory point of unit at sampled point institute envelope is perhaps at the spatial inner obligatory point that adds of sampled point institute envelope, perhaps at the spatial inside and outside obligatory point that adds simultaneously of sampled point institute envelope.
Wherein, the spatial positional information of described obligatory point calculates acquisition according to the spatial positional information of sampled point, and perhaps the mode of selecting by man-machine interactive obtains.
Wherein, if ramose organ is arranged, then the locus of the obligatory point of described branch part obtains by information collecting device.
Wherein, described construction of function and find the solution the unit when constructed fuction is set the functional value of sampled point and obligatory point, and wherein the functional value of sampled point is identical, and the functional value of obligatory point carries out gradual change according to its position in the spatial dimension of sampled point institute envelope.
Wherein, the described construction of function and the function of finding the solution unit structure are formed by one group of RBF linear superposition.
Preferably, described curve reestablishing unit comprises:
Grid is set up module, is used to set up a spatial three-dimensional grid of containing sampled point institute envelope, and calculates the functional value of described function on all grid vertexes;
The implicit surface determination module is used for the difference according to the grid vertex functional value, determine the equivalent implicit surface identical with the sampled point functional value the cell of process;
The contour surface determination module, be used for determined all cells are calculated, solve described equivalent implicit surface and described cell and intersect the one or more triangular facets that form, the final contour surface that obtains by the triangular facet grid representation, i.e. organ curved surface of Chong Jianing.
Wherein, described information collecting device can be gathered sampled point at the inner surface of organ cavity, also can gather sampled point at the outer surface of organ cavity.
Preferably, the quantity of described sampled point is 20~200.
Preferably, the quantity of described obligatory point is 6~70.
Wherein, described information collecting device is electromagnetic-field detector or microwave sounding unit, MSU or acoustic detection device.
Wherein, described electromagnetic-field detector is the conduit that head is equipped with the three-dimensional localization pick off, and wherein said pick off is located based on electromagnetic induction or based on electric field.
Wherein, the three-dimensional position of described catheter head collection can directly be obtained by the three-dimensional localization pick off, also can be obtained by interpolation by two or more three-dimensional localization pick offs.
According to specific embodiment provided by the invention, the invention discloses following technique effect:
The invention provides a kind of method and system of human body organ three-dimensional surface rebuilding, the sparse characteristic of space point set at the collection of tissue lumen surface, obligatory point by replensiher official inside and outside of cavity, and construct a function, make sampled point and obligatory point all satisfy corresponding constraints, solved function obtains function expression then, because all points with same functions value have constituted an implicit surface of this function, extract so carry out contour surface, just can obtain rebuilding curved surface through the cavity of all sampled points according to constraints.
The present invention owing to do not need to construct initial occluding surface, does not have specific (special) requirements to the spatial shape of reconstructed object by constraints solved function expression formula, therefore can accurate reconstruction goes out the complex organ surface model of many bifurcateds.And rebuilding curved surface is space high-order curved surface, and fairing is continuous, meets the characteristic of smooth and continuous under the organ naturalness; And when contour surface extracts,, can obtain very slick surface model, the geometric shape of accurately approaching original cavity according to control to sizing grid.
The specific embodiment
For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, the present invention is further detailed explanation below in conjunction with the drawings and specific embodiments.
The invention provides a kind of method and system of human body organ three-dimensional surface rebuilding, be applicable to and rebuild the sparse organ cavity of sampled point, can accurate reconstruction go out the complex organ surface model of many bifurcateds, and can obtain very slick surface model, the geometric shape of accurately approaching original cavity.The present invention has important application in the minimally invasive surgery of image-guidance, the inner chamber model of structure can be finished at the operation initial stage, and the position of conduit is provided for the doctor in operation process subsequently, improves success rate of operation, reduces operating time.
Specify implementation procedure of the present invention below, still be modeled as example and describe, but the present invention is equally applicable to the reconstruction of the complicated cavity of other single cavitys and many bifurcateds with the chambers of the heart.
With reference to Fig. 1, be the sketch map that chambers of the heart three-dimensional electrolysis is cutd open mapping.
The doctor sends into the inner chamber of heart 2 with a catheter navigation 1 via blood vessel, and at the manipulation in vitro conduit, its stub end 3 is contacted with chambers of the heart inwall 4.Conduit stub end 3 inside include position sensor 5, can detect and be fixed on the magnetic field that outside magnetic field generator produces, and send the information that receives to processor, go out conduit stub end 3 locus of position 6 that contact with wall of the heart by processor according to information calculations with respect to magnetic field generator from pick off.Usually, the doctor can be in operation process control lead 1, gather a plurality of points at inner each position of its interested chambers of the heart, usually more than 50.When gathering these, except the positional information of collection tube stub end 3, also can gather as physiologic informations such as electrocardiosignaies by other functional parts of conduit head end 3.Sampling number is stored in computer according to meeting, processor is according to the position data that comprises in the sampled point, simulate the three-dimensional surface model of mimic chambers of the heart inwall by interpolation algorithm, the electric physiological data by stack corresponding position on surface model obtains the three-dimensional electrolysis of the chambers of the heart and cuts open scaling graph.
With reference to Fig. 2, be the method flow diagram of the described human body organ three-dimensional surface rebuilding of the embodiment of the invention.Easy for describing, be calculated as example with two dimension below and describe, three-dimensional situation can be expanded according to two-dimensional case.
Step 201 is gathered some points according to mode shown in Figure 1, and these points are used to create chambers of the heart surface.All these points all include the coordinate that is collected by the sensor sensing externally-applied magnetic field, have represented the position of conduit head end in real space, and position coordinates has x coordinate, y coordinate and z coordinate.With reference to shown in Figure 3, be the spatial distribution sketch map (with two-dimensional representation) of the data point of being gathered.Usually can gather at least 50 sampled points, the present embodiment hypothesis has ten sampled point S1, S2......S10.Understand for convenience, these sampled points all have been drawn on the two dimensional surface.
Need to prove that described sampled point can be gathered at the organ cavity inner surface, and for the application that needs are gathered organ outer surface point, also can gather at the tissue lumen external surface.Preferably, the quantity of described sampled point is 20~200, can satisfy the needs of algorithm for reconstructing; And according to the function of structure, the quantity of obligatory point is generally about 1/3 of sampled point and promptly can be used for finding the solution, so the quantity of described obligatory point is 6~70.
Step 202, according to the spatial positional information interpolation obligatory point of described sampled point, the value of described obligatory point is as the constraints of function.These obligatory points are used for constructing curve and rebuild function, need choose some as obligatory point cavity inside and outside respectively.The spatial positional information of described obligatory point calculates acquisition according to the spatial positional information of sampled point, and perhaps the mode of selecting by man-machine interactive obtains.The mode that described man-machine interactive is selected is meant by modes of operation such as mouse or keyboards selects to add obligatory point.
This step can be divided into following two sub-steps:
Step a, the step of adding internal constraint point:
Calculate the center of gravity of sampled point space point set, rebuild the obligatory point of curved surface inside as target, the value on the obligatory point is promptly as the constraints of function.With reference to shown in Figure 4,, can draw the geometric center O (x that these put the figure that is surrounded as by coordinate x, y, the z that is had a few asked respectively on average
0, y
0, z
0).
Need to prove, if cavity is a single cavity, then can be with the geometric center that calculates according to step a as an internal constraint point; If cavity is a complicated branch organ, then need be according to mode shown in Figure 1 at organ bifurcation image data point as the internal constraint point, this moment internal constraint point the locus promptly by the sampling acquisition.
Step b, the step of adding external constraint point:
X, y, z coordinate to all sampled points sort, and find out maximum and minima, and outwards expansion, get several and are positioned at the obligatory point of waiting to rebuild the curved surface outside, and the value on the obligatory point is promptly as the constraints of function.
With reference to shown in Figure 4, find out the minima x in the x coordinate of ten sampled points
Min, the maximum x in the x coordinate
Max, the minima Y in the y coordinate
Min, the maximum Y in the y coordinate
Max, the minima z in the z coordinate
Min, the maximum z in the z coordinate
MaxCan define following eight points (Fig. 4 has only listed 4 points of two dimension):
T
1((x
min-d),(y
min-d),(z
min-d)),T
2((x
max+d),(y
min-d),(z
min-d)),
T
3((x
max+d),(y
max+d),(z
min-d)),T
4((x
min-d),(y
max+d),(z
min-d)),
T
5((x
min-d),(y
min-d),(z
max+d)),T
6((x
max+d),(y
min-d),(z
max+d)),
T
7((x
min-d),(y
max+d),(z
max+d)),T
8((x
max+d),(y
max+d),(z
max+d))。
In the following formula, d represents a segment length, can get 10mm, also can get other numerical value.Cuboid with an extension of this eight somes formations has surrounded all sampled points.
Need to prove that the peripheral shape of surrounding all sampled points can have multiple arbitrary form, not necessarily cuboid is also not necessarily got its 8 end points, as long as these all sampled points of geometry physical ability envelope, the present invention does not limit at this.
Step 203 is constructed a function, makes described sampled point and obligatory point all satisfy corresponding constraints.
In the present embodiment, having made up one is the function f of parameter with the space coordinates, and certain has following form on a bit to f in the space:
Wherein, φ (x) is a RBF, and the RBF of plurality of optional is arranged, preferably φ (x)=| x|
3, ω is the weight of each RBF correspondence.A is the coordinate of this point, at sampled point S
iOn,
Then with φ
IjRepresent S
jPoint is to S
iThe RBF of effect, φ
Ij=φ (S
i-S
j),
As shown from the above formula, this function is formed by one group of RBF linear superposition.
In addition, when constructed fuction, set the functional value of sampled point and obligatory point, wherein the functional value of sampled point is identical, and the functional value of obligatory point carries out gradual change according to its position in the spatial dimension of sampled point institute envelope.For example, the functional value that regulation internal constraint point O is ordered is 1, external constraint point T
1, T
2, T
3, T
4, T
5, T
6, T
7, T
8Functional value be-1, the functional value on all sampled points is 0.Bring the functional value of these points into this function as constraints, obtaining one is the system of linear equations of independent variable with ω, as follows:
The above-mentioned f of prescribed function artificially gets the 1,0, the-the 1st at the inside and outside functional value opposite in sign of three-dimensional surface in the present embodiment, for convenience of calculation, can certainly get other numerical value.
Function f also can have other expression formula, for example the linear term in the formula removed,
RBF φ (x) wherein also can be other form, as φ (x)=x
2
Step 204 is found the solution this function by described constraints, obtains the expression formula of this function.
By sophisticated algorithm this system of linear equations is carried out numerical solution and draw one group of ω and P, and then draw the expression formula of f.Above-mentioned function definition a plurality of implicit surfaces, all points with same functions value have constituted an implicit surface of this function.In whole space, all functional values equal the curved surface that 0 point constitutes a sealing, promptly satisfy the f implicit surface of expression formula f=0.All sampled points are all on this implicit surface, and this implicit surface is the cavity curved surface of being asked that all sampled point interpolation are drawn.
By each parameter of constraints solved function expression formula, be a calculus of variations problem, promptly adopt this function of calculus of variations numerical solution.Because it is very ripe to find the solution the method for system of linear equations at present, the present invention does not elaborate at this.
By above step, only obtained describing the function of this curved surface, also need to carry out contour surface and extract according to constraints, obtain rebuilding curved surface through the cavity of all sampled points.Described contour surface extracts the implicit surface that promptly extracts f=0, specifically comprises following several steps:
Step 205 is set up the three-dimensional grid of an envelope cavity curved surface, and calculates the functional value of described function on all lattice points.
With reference to Fig. 5, be the grid sketch map of the envelope cavity curved surface of foundation.Is several equal-sized cubic units lattice with the space along the change in coordinate axis direction orthogonal division, forms a three-dimensional grid.Among the figure, cube has been simplified shown as the square on the plane.The length of side of cell is a less numerical value, can be set up on their own by user, and the length of side of cell is more for a short time usually means that final simulation curved surface is smooth more.Then, computer calculates the functional value of function f on each lattice point according to the position coordinates of each lattice point on the grid.
Step 206, according to the functional value and the constraints of lattice point, choose the cavity curved surface the cell of process.
With reference to Fig. 6, be the cavity curved surface the cell sketch map of process.The value symbol of predefine curved surface internal constraint point and external constraint point antithesis in the present embodiment, can prove on the mathematics, if the sign of the functional value on eight summits of certain cell is inconsistent, promptly existing on the occasion of negative value is arranged again, functional value is that 0 implicit surface must pass through this cell.Therefore, by calculating the functional value on all grid lattice points, and utilize described constraints, just can find out and satisfy at its eight inconsistent all cells of lattice point superior function value sign symbol, these cells are the cell that passes through the cavity curved surface.
At this moment, what obtain is an approximate curved surface of being represented by the cubic units lattice, if accurately simulate the cavity curved surface, also needs described approximate curved surface is handled.
Step 207 is calculated each cell that passes through curved surface, draws out by the cavity of triangular facet grid representation and rebuilds curved surface.
With reference to Fig. 7, be to pass through the sketch map that cell is drawn the cavity curved surface.Each selected cell lattice is inevitable adjacent with several cells that satisfy condition equally.Computer can be by the algorithm that sets in advance, and according to each cell and the adjacent situation of cell on every side, draws a cross section of passing through this cell in each cell.Because the internal point of the curved surface of asking and the functional value opposite in sign of external point, so sign according to each summit of cell, according to the distribution of contour surface and each bar limit intersection point of cell, can express the cross section of intersecting of contour surface and cell with one or more triangular facets.
Contour surface is all set with certain same alike result value spatial point.If a bit (scalar attribute value z) is designated as f, and (z), contour surface can be expressed as 3 d data field so for x, y for x, y
{(x,y,z)|f(x,y,z)=magic}
Wherein magic is the attribute threshold value of setting.Contour surface is to embed three-dimensional curved surface, when curved surface and certain voxel cell lattice intersect, cross surface is one or more patch, the essential idea of Marching Cubes algorithm (hereinafter to be referred as the MC algorithm) is that the intersection point with contour surface and hexahedron voxel limit connects into closed polygon, approaches that curved surface and voxel intersect and the patch that forms with this polygon.Be not that whole voxels in the 3 d-dem data fields all intersect with contour surface, when the property value on eight summits of voxel all greater than or during all less than magic, this voxel and contour surface do not have cross surface; Have only those existing summits that the dough sheet that approaches that just contains contour surface less than the voxel on the summit of magic is arranged again, claim that such voxel is a boundary voxel greater than magic.Figure 10 lists boundary voxel and the crossing polygonal mesh generating mode of contour surface.
Step 208, final, all triangular facets have been spliced to form the three-dimensional surface of a sealing, are the three-dimensional reconstruction curved surface of being asked.With reference to Fig. 8, be the sketch map of the cavity curved surface that is spliced to form by all triangular facets.
Because the reconstruction curved surface that generates is not considered the geometric shape of original curved surface according to the method described above, the triangular facet mesh-density only depends on the density of space cube grid, triangular facet enormous amount after all are rebuild influences that follow-up curved surface shows and the efficient of analytical calculation.So preferably, the present invention adopts algorithm that the triangular facet grid simplifies to reduce triangular facet grid number in the reconstruction model.
The essence of graticule model simplification is: under the situation that keeps the archetype feature as far as possible, reduce the triangular facet of archetype and the number on summit to greatest extent.In numerous shortcut calculations, representative practical algorithm has the 1. vertex deletion algorithm of Schroeder; 2. the Garland summit is to shrinkage method; 3. the progressive grid of Hoppe (Progressive Mesh) thinning method.
The Schroeder shortcut calculation at first utilizes the local geometric on each summit and topology information with each vertex classification, and the judgment criteria according to different summits determines whether this summit can delete then.If can delete, then adopt recurrence method that triangulation is carried out in left cavity after deleting the summit, otherwise keep this summit.This method is specially adapted to the lattice simplified calculating of triangular facet that Marching Cubes algorithm obtains, and the vertex set of simplifying Vee formation veil lattice is the subclass of original mesh vertex set, does not produce new summit in the simplification process, does not also change the position of original vertices.The Schroeder shortcut calculation keeps the topological structure of original mesh, this point is extremely important to medical images destination organization three-dimensional reconstruction, to produce new problem be the triangular facet that has many " long and narrow " in the model after simplifying but do not change the original vertices geometric position, and the geometric shape of grid is relatively poor.
The Garland algorithm is the method for simplifying that the summit is combined.Two summits of summit centering are not necessarily on same limit on the model.The Garland algorithm thinks that each summit is the intersection point that comprises whole triangular facets on this summit in the grid, to each summit determine one with each triangular facet method vector and the relevant quadratic matrix Q in locus, the vertical dimension quadratic sum of each triangular facet of space any point v under to this summit is v
TQv.When a pair of vertex v
i, v
jAfter the merging, obtain a new summit
If this a pair of vertex v
i, v
jThe corresponding quadratic matrix Q of difference
1And Q
2, the summit
Replace with these two errors that the summit produced and be respectively
With
Garland will
Be defined as and merge the summit v
i, v
jThe error cost.Set up according to the error cost and can merge a little right Priority Queues (Priority Queue), from formation point of every merging to after, the renewal Priority Queues.Adopt Priority Queues, the point of each pooled error cost minimum is right, up to the final requirement of satisfying the triangular facet number.Adopt complete binary tree heap construction Priority Queues, lower, the lattice simplified fast operation of time complexity of the insertion of formation, deletion and renewal operation.The Garland algorithm can change the geometric position of grid original vertices, change the topological structure of original mesh, and can punish by the point that causes " long and narrow " triangular facet is combined, thereby the triangular facet geometric shape of the grid that produces is more excellent.Though the Garland algorithm changes the topological structure of grid, it can carry out step-feeding to grid and simplify, and helps the multi-resolution representation and the network transmission of model.
The present invention adopts the Garland algorithm that grid is simplified, improve error calculation method simultaneously, keep characteristic edge in the grid, the topological structure that keeps original mesh, and be principle with the error minimum, set to simplify the summit of back model and the ratio of archetype number of vertices, preferably select 25%~75%, make error minimum between simplified model and the archetype.The number of the final grid intermediate cam face that obtains depends on the geometric shape of curved surface itself, and distributing with original sample point has nothing to do, the zone that curvature of curved surface is big, and veil lattice density is big, and the flat site veil lattice density that curvature is little is low.With reference to Fig. 8, be the sketch map of the cavity curved surface that is spliced to form by all triangular facets.
In the above-mentioned process of reconstruction, can control sizing grid when extracting, so can obtain very slick surface model, the geometric shape of accurately approaching original cavity owing to contour surface.And rebuilding curved surface is space high-order curved surface, and fairing is continuous, meets the characteristic of smooth and continuous under the organ naturalness.
The present invention is satisfying under the smooth mimic condition, can also rebuild the complex organ housing surface of multiple-limb.Because the present invention is by constraints solved function expression formula, do not need to construct initial occluding surface, spatial shape to reconstructed object does not have specific (special) requirements, and by adding obligatory point in the cavity bifurcation, can accurate reconstruction goes out the complex organ surface model of many bifurcateds.
The present invention also provides a kind of system embodiment of human body organ three-dimensional surface rebuilding, with reference to Fig. 9, is the system construction drawing of described human body organ three-dimensional surface rebuilding.Described system mainly comprises information collecting device 91 and information processor 92 two parts, wherein information collecting device 91 is responsible for gathering from the tissue lumen surface spatial positional information of several sampled points, and information processor 92 is responsible for utilizing these sampling point informations to simulate tissue lumen surface model.Specify as follows:
In actual applications, information collecting device 91 is realized by catheter navigation shown in Figure 1 usually, catheter head is equipped with the three-dimensional localization pick off, imports human body and gathers three dimensional local information at inner chamber or the short stay of organ surface place, and send positional information to information processor 92.Catheter head can be installed one or more three-dimensional localization pick offs, and pick off can be located based on electromagnetic induction, also can locate based on electric field, can also adopt other locate modes.And the three-dimensional position of catheter head collection can directly be obtained by the three-dimensional localization pick off, also can be obtained by interpolation by two or more three-dimensional localization pick offs.
Information processor 92 mainly comprises construction of function and finds the solution unit 921 and curve reestablishing unit 922.Wherein, construction of function and find the solution unit 921 and be responsible for adding obligatory points according to the spatial positional informations of described sampled point, and construct a function, make described sampled point and obligatory point all satisfy corresponding constraints.These obligatory points comprise the internal constraint point and the external constraint point of cavity curved surface, and the value of obligatory point is mainly used in constructing curve and rebuilds function as the constraints of function.Wherein, the external constraint point utilizes the sampling point position information of housing surface to calculate and can obtain, and internal constraint point also can obtain by calculating usually, but if the complicated branch cavity then needs to obtain the positional information of branch part as the internal constraint point by sampling.The external constraint point need all be surrounded reconstructed object, and the peripheral shape of surrounding all sampled points also can have multiple arbitrary form.The detailed process of adding obligatory point is with reference to aforementioned.
Construction of function and the function of finding the solution unit 921 structure are formed by one group of RBF linear superposition, and for example wherein a kind of expression-form is:
Function f also can have other expression formula.
Carry out assignment by the sampled point to inner obligatory point, external constraint point and housing surface, for example value is 1 ,-1,0 respectively, and obtaining one is the system of linear equations of independent variable with ω, as previously shown.
Construction of function and find the solution unit 921 and also be responsible for described system of linear equations is found the solution carries out numerical solution by sophisticated algorithm to this system of linear equations and draws one group of ω and P, and then draws the expression formula of f.One group of implicit surface of this function representation, wherein all points with same functions value have constituted an implicit surface of this function, and the implicit surface of f=0 is the cavity curved surface of all sampled points of process of being asked.
Curve reestablishing unit 922 is responsible for according to this function expression and constraints, reconstructs through the cavity of all sampled points and rebuilds curved surface.In the present embodiment, the main methods that adopt contour surfaces to extract in curve reestablishing unit 922 obtain final triangular facet grid models, and curve reestablishing unit 922 comprises that further grid sets up module a, implicit surface determination module b and contour surface determination module c.
Wherein, grid is set up the three-dimensional grid that module a is responsible for setting up an envelope cavity curved surface, and calculates the functional value of described function on all lattice points, with reference to shown in Figure 5.
Implicit surface determination module b be responsible in described three-dimensional grid, determining the cavity curved surface the cell of process, with reference to shown in Figure 6.Can prove on the mathematics, the cavity curved surface the cell of process, sign of the functional value on each summit is inconsistent on it.Because the value symbol of predefine curved surface internal constraint point and external constraint point antithesis, by calculating the functional value on all grid lattice points, and utilize described constraints, just can find out and satisfy at its eight inconsistent all cells of lattice point superior function value sign symbol, these cells are the cell that passes through the cavity curved surface.
Contour surface determination module c is responsible for each cell that passes through curved surface is calculated, and draws out by the cavity of triangular facet grid representation and rebuilds curved surface.When drawing curved surface, according to the relative position relation between cell and the adjacent some cells that satisfy similarity condition, draw out the triangular facet of transversal this cell in this cell inside on every side, reference is shown in Figure 7.Finally, all triangular facets have been spliced to form the three-dimensional surface of a sealing, are the three-dimensional reconstruction curved surface of being asked, with reference to shown in Figure 8.
Housing surface reconstructing system shown in Figure 9 at the sparse characteristic of sampled point, by add obligatory point in the cavity bifurcation, can be rebuild the complex organ housing surface of multiple-limb; And, can obtain very slick surface model, the geometric shape of accurately approaching original cavity by controlling the cell size of three-dimensional grid.
The part that does not describe in detail in the system shown in Figure 9 can be considered for length referring to the relevant portion of Fig. 1-method shown in Figure 8, is not described in detail in this.
More than to the method and system of a kind of human body organ three-dimensional surface rebuilding provided by the present invention, be described in detail, used specific case herein principle of the present invention and embodiment are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, part in specific embodiments and applications all can change.In sum, this description should not be construed as limitation of the present invention.