CN109859322A - A kind of spectrum posture moving method based on deformation pattern - Google Patents

A kind of spectrum posture moving method based on deformation pattern Download PDF

Info

Publication number
CN109859322A
CN109859322A CN201910056771.5A CN201910056771A CN109859322A CN 109859322 A CN109859322 A CN 109859322A CN 201910056771 A CN201910056771 A CN 201910056771A CN 109859322 A CN109859322 A CN 109859322A
Authority
CN
China
Prior art keywords
grid
deformation
pose
source
vertex
Prior art date
Legal status (The legal status 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 status listed.)
Granted
Application number
CN201910056771.5A
Other languages
Chinese (zh)
Other versions
CN109859322B (en
Inventor
尹梦晓
苏鹏
林振峰
杨锋
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangxi University
Original Assignee
Guangxi University
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 Guangxi University filed Critical Guangxi University
Priority to CN201910056771.5A priority Critical patent/CN109859322B/en
Publication of CN109859322A publication Critical patent/CN109859322A/en
Application granted granted Critical
Publication of CN109859322B publication Critical patent/CN109859322B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Processing Or Creating Images (AREA)
  • Prostheses (AREA)

Abstract

本发明公开了一种基于变形图的谱姿态迁移方法，包括步骤：1)利用网格简化算法简化源网格生成源网格变形图；2)利用参考网格对源网格进行基于耦合准调和基的谱姿态迁移；3)根据姿态迁移结果和源网格变形图利用最优化能量函数生成姿态迁移后变形图；4)根据姿态迁移后生成的变形图利用嵌入变形图编辑方法变形源网格生成目标网格；5)分割所有网格姿态迁移不充分区域，进行分层姿态迁移，直到姿态迁移充分为止。本发明利用嵌入变形图编辑方法表示源网格时具有较好的局部性，在一定程度上降低了几何代理质量对结果的影响；对子网格使用简单偏移拼接，不必进行朝向调整，有效提高姿态迁移生成目标网格的质量。

The invention discloses a method for migrating spectral attitude based on deformation graph, comprising the steps of: 1) simplifying a source grid by a grid simplification algorithm to generate a source grid deformation graph; 2) using a reference grid to perform coupling-based calibration on the source grid The spectral pose transfer of the harmonic basis; 3) According to the pose transfer result and the source mesh deformation map, use the optimized energy function to generate the post-pose-transfer deformation map; 4) According to the deformation map generated after the pose transfer, use the embedded deformation map editing method to deform the source mesh 5) Divide all the grids with insufficient pose transfer area, and perform hierarchical pose transfer until the pose transfer is sufficient. The invention uses the embedded deformation map editing method to represent the source grid, and has better locality, which reduces the influence of geometric proxy quality on the result to a certain extent; uses simple offset splicing for sub-grids, and does not need to adjust the orientation, effectively Improve the quality of the target mesh generated by pose transfer.

Description

A kind of spectrum posture moving method based on deformation pattern
Technical field
The present invention relates to the technical fields of 3D model spectra posture migration, refer in particular to a kind of spectrum posture based on deformation pattern and move Shifting method.
Background technique
Three-dimensional grid model 3D printing, virtual reality, in terms of be widely used.It is built using traditional-handwork Mould, scanning device or modeling software are difficult rapidly and accurately to model complicated geometry.And to existing grid mould Type is edited and is reused, can be to avoid modeling again.It is moved as the mesh modeling technology deformation migration based on sample with posture It moves and obtains the target gridding with similar posture using the existing posture of the grid of reference but how accurately to describe source grid mould Type posture, automatic guiding target grid model deformation is also a challenging project.
Characteristics of low-frequency function (the i.e. smaller characteristic value that L é vy passes through Laplce (Laplacian) matrix of exchange grid Corresponding characteristic function) corresponding coefficient carries out ordinary posture migration between the grid with identical connection relationship.But Due to the difference of source grid and Laplce's feature base of the grid of reference, its expression of results is caused notable difference occur.Posture is moved The target gridding model obtained after shifting will appear serious distortion and metaboly, and posture study is not enough.Kovnatsky Deng the Laplacian Matrix feature base of the different grid model of two connection relationships of optimization, obtain based on Functional Mapping have it is simultaneous The coupling quasi- reconciliation base of capacitive, then it is different to be attached relationship for the quasi- low frequency coefficient for reconciling base of the coupling of two grids of exchange , the migration of posture between the grid model that posture is different.Although this method solve due to the different caused shape of feature basis Twisted phenomena, but posture study is still not enough.Yin etc. proposes the spectrum posture migration algorithm of details holding and multilayer is moved Move frame.This method is on the basis of reconciliation base posture migration quasi- based on coupling, in conjunction with the subspace based on broad sense centre coordinate Technology, being acted on behalf of with Cage as geometry reduces solution scale, reduces the freedom degree of deformation, guarantees the stability solved.To understand Certainly posture learns insufficient problem, and the propositions such as Yin are layered posture migration strategy, and the region insufficient to migration posture is divided It cuts, will not be that the posture of large scale is converted into the large scale posture of regional area, then carries out spectrum posture migration originally.To obtain Preferable posture migration effect.But since this method indicates source grid using Cage and HCCI combustion, and HCCI combustion is discontented The internal locality of foot, therefore posture migration results are affected by Cage.
The present invention provides a kind of spectrum posture moving method based on deformation pattern, is replaced using embedding distortion figure edit methods wide The sub-space technique of adopted centre coordinate, using deformation pattern as geometry, agency, geodesic distance indicate source grid as weight.Embedding distortion The geometric detail of grid can be preferably kept, this reduces influence of the geometry agency to migration results to a certain extent, to mention The quality of high migration results.
Summary of the invention
The purpose of the present invention is to overcome the shortcomings of the existing technology and deficiency, proposes a kind of spectrum posture based on deformation pattern Moving method simplifies source grid using optimization second order error measure and obtains deformation pattern, recycles embedding distortion figure editing side Method replaces the sub-space technique of broad sense centre coordinate, and using deformation pattern as geometry, agency, geodesic distance indicate source grid as weight. Embedding distortion can preferably keep the geometric detail of grid, this reduces geometry agency to the shadow of migration results to a certain extent It rings, to improve the quality of migration results.
To achieve the above object, a kind of technical solution provided by the present invention are as follows: spectrum posture migration side based on deformation pattern Method, comprising the following steps:
1) simplify source grid using Mesh simplification algorithm and generate source distortion of the mesh figure；
2) source grid migrated based on the quasi- spectrum posture for reconciling base of coupling using the grid of reference；
3) it is desired to make money or profit according to posture migration results and source distortion of the mesh and is deformed after generating posture migration with optimized energy function Figure；
4) deformation pattern generated after being migrated according to posture generates target using embedding distortion figure edit methods deformation sources grid Grid；
5) divided ownership grid posture migrates insufficient region, carries out layering posture migration according to step 2) to step 4), Until posture migration is abundant.
In step 1), optimization second order error metric algorithm, the i.e. letter of QEM (quadric error metric) algorithm are utilized Change source grid generates deformation pattern, optimizes total collapse cost of grid each edge in second order error metric algorithm is defined as:
Cost (a, b)=α QQEM(a,b)+βQF(a,b)+γQReg(a,b)+μQAre(a,b)
In formula, a, b are two endpoints on side to be folded respectively, and α, β, γ, μ are respectively scalar weight, QQEMFor the side QEM folding Iterate valence, QFIt is the offset on adjacent smallest offset degree side after edge contraction to be folded, QRegWhen for folded edge (a, b) → v just Then property cost, QReg=| Reg (NF(v))-max{Reg(NF(a)),Reg(NF(b)) } |, NF(a)、NF(b)、NFBe (v) vertex a, B, the adjoining triangle sets of v, Reg (tri)=3-2R (tri), R (tri)=cos (∠ α)+cos (∠ β)+cos (∠ γ), For the regularity of Δ tri,Wherein S (tri) is the area of triangle tri.
In step 2), source grid migrated based on the quasi- spectrum posture for reconciling base of coupling using the grid of reference, including Following steps:
2.1) input grid of reference M, the Laplacian Matrix of source grid M ' and its corresponding spectrum and characteristic function, meter are calculated Calculate the Laplacian Matrix of deformation pattern G；
2.2) grid of reference and the direct feature corresponding points of source grid are chosen, the base between the grid of reference and source grid is calculated In area approximation s neighborhood point of corresponding points indicator function as the respective function between grid；
2.3) the Laplce's feature base for optimizing grid of reference M and source grid M ' obtains coupling the quasi- base { φ that reconcilesi M}、 {φi M', utilize formulaFind out the posture migration results based on the quasi- reconciliation base of coupling In formula, αi MAnd αi M'The respectively spectral coefficient of M and M ', k are exchange αi M'Number, | V'| be M ' number of vertices.
In step 3), after carrying out posture migration using the quasi- reconciliation base of coupling, posture migration results is utilized and optimize energy Flow functionFind out the deformation pattern after posture migrates, point constraint item in formulaProtect Laplce's coordinate itemScale bound termQ is geodesic distance weight matrix,For the deformation pattern solved,For deformation pattern to be solved Vertex set,ForNumber of vertices,For based on coupling it is quasi- reconcile base carry out posture migration as a result,For rotation Turn transformation matrix, δi GFor Laplce's coordinate of deformation pattern G,By solution deformation pattern Laplce's coordinate,ForScaled matrix, λ1、λ2For scalar weight.
In step 4), obtained deformation pattern will be solved and generate target using embedding distortion figure edit methods deformation sources grid Grid, comprising the following steps:
4.1) minimum energy function is utilizedOptimize posture Migration results obtain the rotational translation matrix of source distortion of the mesh, whereinForNumber of vertices, rotate itemSingle vertex rotates item Rot (Rj)=(c1·c2)2+(c1·c3)2+(c2·c3)2+(c1·c2)2+ (c1·c1-1)2+(c2·c2-1)2+(c3·c3-1)2,c1、c2And c3It is deformation pattern apex spin matrix RjColumn vector, rule Then change itemN (j) is vertex vjNeighborhood, t be translation Transformation,For deformation pattern vertex, αjn=1.0, point constraint item For the target gridding of generation On vertex,For deformation patternVertex, wrot、wreg、wconFor scalar weight；
4.2) formula is utilized after obtaining rotational translation matrix Calculate target griddingVertex setWherein wj(vl') it is geodesic distance weight,For spin matrix,It is flat Move matrix, vl' be source grid vertex, gj(vi') be and source grid vertex vi' on m nearest deformation pattern G of geodesic distance Vertex set,For the vertex on required final goal grid.
In step 5), the segmentation grid of reference, the target gridding of generation and its deformation pattern posture migrate insufficient region, root Layering posture migration is carried out according to step 2) to step 4), until posture migration is abundant, is included the following steps:
5.1) choose posture learn insufficient Local grid, by corresponding grid of reference M, generation target gridding M ', Deformation patternIt is split, obtains corresponding local subnet lattice S, S ' and Gs；
5.2) local deformation figure G is found out using step 2) to step 3)sDeformation pattern after posture migration
5.3) G is calculatedsMean deviation amount of the partitioning boundary vertex position before and after posture migration, is added toPhase Answer vertex；
5.4) it utilizesVertex information update deformation pattern
5.5) final carriage migration results are obtained using step 4).
Compared with prior art, the present invention have the following advantages that with the utility model has the advantages that
1, the present invention simplifies source grid using optimization second order error measurement (quadric error metric, QEM) algorithm It generates vertex to be evenly distributed deformation pattern, better effect can be generated in driving source distortion of the mesh.
2, the present invention using embedding distortion edit methods replace broad sense centre coordinate sub-space technique, using deformation pattern as Geometry agency, geodesic distance weight indicate source grid, effectively improve posture migration and generate target gridding quality.
3, when being layered posture migration, directly translation regional area grid (sub-grid) carries out and integral grid the present invention Splicing, it is not necessary to carry out towards adjustment.
Detailed description of the invention
Fig. 1 is that posture of the present invention migrates flow chart.
Fig. 2 is that posture of the present invention migrates flow diagram.
Fig. 3 is that the present invention carries out carrying out posture migration results schematic diagram based on the quasi- base that reconciles of coupling.
Fig. 4 is that posture of the present invention migrates deformation result diagram intention.
Specific embodiment
The present invention is further explained in the light of specific embodiments.
As depicted in figs. 1 and 2, the spectrum posture moving method based on deformation pattern provided by the present embodiment, input grid and Source grid comprising following steps:
1) simplify source grid using optimization second order error measurement (quadric error metric, QEM) algorithm and generate top The deformation pattern that point is evenly distributed optimizes total collapse cost of grid each edge in second order error metric algorithm is defined as:
Cost (a, b)=α QQEM(a,b)+βQF(a,b)+γQReg(a,b)+μQAre(a,b)
In formula, a, b are two endpoints on side to be folded respectively, and α, β, γ, μ are respectively scalar weight, QQEMFor the side QEM folding Iterate valence, QFIt is the offset on adjacent smallest offset degree side after edge contraction to be folded, QRegWhen for folded edge (a, b) → v just Then property cost, QReg=| Reg (NF(v))-max{Reg(NF(a)),Reg(NF(b)) } |, NF(a)、NF(b)、NFBe (v) vertex a, B, the adjoining triangle sets of v, Reg (tri)=3-2R (tri), R (tri)=cos (∠ α)+cos (∠ β)+cos (∠ γ), For the regularity of Δ tri,Wherein S (tri) is the area of triangle tri.
2) source grid migrated based on the quasi- spectrum posture for reconciling base of coupling using the grid of reference, comprising the following steps:
2.1) input grid of reference M, the Laplacian Matrix of source grid M ' and its corresponding spectrum and characteristic function, meter are calculated Calculate the Laplacian Matrix of deformation pattern G；
2.2) grid of reference and the direct feature corresponding points of source grid are chosen, the base between the grid of reference and source grid is calculated In area approximation s neighborhood point of corresponding points indicator function as the respective function between grid；
2.3) the Laplce's feature base for optimizing grid of reference M and source grid M ' obtains coupling the quasi- base { φ that reconcilesi M}、 {φi M', utilize formulaFind out the posture migration results based on the quasi- reconciliation base of coupling In formula, αi MAnd αi M'The respectively spectral coefficient of M and M ', k are exchange αi M'Number, | V'| be M ' number of vertices.The base of generation It is as shown in Figure 3 in the spectrum posture migration results of the quasi- reconciliation base of coupling.
3) after carrying out posture migration using the quasi- reconciliation base of coupling, posture migration results are utilized into optimized energy functionDeformation pattern after finding out posture migration, as shown in figure 4, point constraint item in formulaProtect Laplce's coordinate itemScale bound termQ is geodesic distance weight matrix,For the deformation pattern solved,For deformation pattern to be solved Vertex set,ForNumber of vertices,For based on coupling it is quasi- reconcile base carry out posture migration as a result,For rotation Turn transformation matrix, δi GFor Laplce's coordinate of deformation pattern G,By solution deformation pattern Laplce's coordinate,ForScaled matrix, λ1、λ2For scalar weight.
4) obtained deformation pattern will be solved and generates target gridding using embedding distortion figure edit methods deformation sources grid, including Following steps:
4.1) minimum energy function is utilizedOptimize posture Migration results obtain the rotational translation matrix of source distortion of the mesh, whereinForNumber of vertices, rotate itemSingle vertex rotates item Rot (Rj)=(c1·c2)2+(c1·c3)2+(c2·c3)2+(c1·c2)2+ (c1·c1-1)2+(c2·c2-1)2+(c3·c3-1)2,c1、c2And c3It is deformation pattern apex spin matrix RjColumn vector, rule Then change itemN (j) is vertex vjNeighborhood, t be translation Transformation,For deformation pattern vertex, αjn=1.0, point constraint item For the target gridding of generation On vertex,For deformation patternVertex, wrot、wreg、wconFor scalar weight；
4.2) formula is utilized after obtaining rotational translation matrix Calculate target griddingVertex setWherein wj(vl') it is geodesic distance weight,For spin matrix,It is flat Move matrix, vl' be source grid vertex, gj(vi') be and source grid vertex vi' on m nearest deformation pattern G of geodesic distance Vertex set,For the vertex on required final goal grid.
The spectrum posture migration low frequency posture migration results based on deformation pattern generated are as shown in the step 2-4 in Fig. 2.
5) divide the grid of reference, the target gridding of generation and its deformation pattern posture and migrate insufficient region, according to step 2) Layering posture migration is carried out to step 4), until posture migration is abundant, is included the following steps:
5.1) choose posture learn insufficient Local grid, by corresponding grid of reference M, generation target gridding M ', Deformation patternIt is split, obtains corresponding local subnet lattice S, S ' and Gs；
5.2) local deformation figure G is found out using step 2) to step 3)sDeformation pattern after posture migration
5.3) G is calculatedsMean deviation amount of the partitioning boundary vertex position before and after posture migration, is added toPhase Answer vertex；
5.4) it utilizesVertex information update deformation pattern
5.5) final carriage migration results are obtained using step 4).
Posture transition process and final result are layered as shown in the step 5 in Fig. 2.
In conclusion the present invention provides new method for three-dimensional model attitude migration after using above scheme, utilize Embedding distortion figure edit methods indicate there is preferable locality when the grid of source, reduce geometry agent quality to a certain extent Influence to result；Simple offset splicing is used to sub-grid, it is not necessary to carry out towards adjustment.It effectively improves posture migration and generates mesh The quality of grid is marked, there is actual promotional value, be worthy to be popularized.
Embodiment described above is only the preferred embodiments of the invention, and but not intended to limit the scope of the present invention, therefore All shapes according to the present invention change made by principle, should all be included within the scope of protection of the present invention.

Claims (6)

1.一种基于变形图的谱姿态迁移方法，其特征在于，包括以下步骤：1. a kind of spectral pose migration method based on deformation map, is characterized in that, comprises the following steps: 1)利用网格简化算法简化源网格生成源网格变形图；1) Simplify the source mesh by using the mesh simplification algorithm to generate the source mesh deformation map; 2)利用参考网格对源网格进行基于耦合准调和基的谱姿态迁移；2) Using the reference grid to perform spectral pose transfer based on the coupled quasi-harmonic basis on the source grid; 3)根据姿态迁移结果和源网格变形图利用最优化能量函数生成姿态迁移后变形图；3) According to the attitude transfer result and the source mesh deformation map, the optimized energy function is used to generate the deformation map after the attitude migration; 4)根据姿态迁移后生成的变形图利用嵌入变形图编辑方法变形源网格生成目标网格；4) According to the deformation map generated after the pose migration, using the embedded deformation map editing method to deform the source mesh to generate the target mesh; 5)分割所有网格姿态迁移不充分区域，根据步骤2)至步骤4)进行分层姿态迁移，直到姿态迁移充分为止。5) Divide all grid pose transfer insufficient regions, and perform hierarchical pose transfer according to step 2) to step 4) until the pose transfer is sufficient. 2.根据权利要求1所述的一种基于变形图的谱姿态迁移方法，其特征在于：在步骤1)中，利用优化二次误差度量算法，即QEM算法简化源网格生成变形图，优化二次误差度量算法中网格每条边的总折叠代价定义为:2. a kind of spectral attitude migration method based on deformation map according to claim 1, is characterized in that: in step 1), utilize optimization secondary error measurement algorithm, namely QEM algorithm simplifies source grid to generate deformation map, optimize The total folding cost of each edge of the grid in the quadratic error metric algorithm is defined as: cost(a,b)＝αQQEM(a,b)+βQF(a,b)+γQReg(a,b)+μQAre(a,b)cost(a,b)=αQ QEM (a,b)+βQ F (a,b)+γQ Reg (a,b)+μQ Are (a,b) 式中，a、b分别是待折叠边的两个端点，α、β、γ、μ分别为标量权值，QQEM为QEM边折叠代价，QF是待折叠边折叠后相邻最小偏移程度边的偏移量，QReg为折叠边(a,b)→v时的正则性代价，QReg＝|Reg(NF(v))-max{Reg(NF(a)),Reg(NF(b))}|，NF(a)、NF(b)、NF(v)为顶点a、b、v的邻接三角形集合，Reg(tri)＝3-2R(tri)，R(tri)＝cos(∠α)+cos(∠β)+cos(∠γ)，为Δtri的正则性，其中S(tri)为三角形tri的面积。In the formula, a and b are the two endpoints of the edge to be folded, α, β, γ, and μ are the scalar weights, respectively, Q QEM is the QEM edge folding cost, and Q F is the minimum adjacent offset after the edge to be folded is folded. Offset of degree edge, Q Reg is the regularity cost when folding edge (a,b)→v, Q Reg =|Reg(N F (v))-max{Reg(N F (a)),Reg ( NF (b))}|, NF(a), NF (b), NF (v) are adjacent triangle sets of vertices a, b, v, Reg(tri)= 3-2R (tri) , R(tri)=cos(∠α)+cos(∠β)+cos(∠γ), which is the regularity of Δtri, where S(tri) is the area of the triangle tri. 3.根据权利要求1所述的一种基于变形图的谱姿态迁移方法，其特征在于：在步骤2)中，利用参考网格对源网格进行基于耦合准调和基的谱姿态迁移，包括以下步骤：3. a kind of spectral attitude migration method based on deformation map according to claim 1, is characterized in that: in step 2), utilize reference grid to carry out the spectral attitude migration based on coupling quasi-harmonic basis to source grid, comprising The following steps: 2.1)计算输入参考网格M、源网格M′的拉普拉斯矩阵及其对应的谱与特征函数，计算变形图G的拉普拉斯矩阵；2.1) Calculate the Laplacian matrix of the input reference grid M, the source grid M' and its corresponding spectrum and eigenfunction, and calculate the Laplacian matrix of the deformation graph G; 2.2)选取参考网格与源网格直接的特征对应点，计算参考网格与源网格之间的基于面积近似的对应点s个邻域点的指示函数作为网格之间的对应函数；2.2) Select the feature corresponding points directly between the reference grid and the source grid, and calculate the indicator functions of the corresponding points s neighborhood points based on the area approximation between the reference grid and the source grid as the corresponding function between the grids; 2.3)优化参考网格M和源网格M′的拉普拉斯特征基，得到耦合准调和基{φi M}、{φi M'}，利用公式求出基于耦合准调和基的姿态迁移结果式中，αi M和αi M'分别为M和M′的谱系数，k为交换αi M'的个数，|V'|为M′的顶点个数。2.3) Optimize the Laplacian eigenbases of the reference grid M and the source grid M′ to obtain the coupled quasi-harmonic basis {φ i M }, {φ i M' }, using the formula Obtain the attitude transfer result based on the coupled quasi-harmonic basis In the formula, α i M and α i M' are the spectral coefficients of M and M' respectively, k is the number of exchange α i M' , |V'| is the number of vertices of M'. 4.根据权利要求1所述的一种基于变形图的谱姿态迁移方法，其特征在于：在步骤3)中，利用耦合准调和基进行姿态迁移后，将姿态迁移结果利用最优化能量函数求出姿态迁移后的变形图，式中顶点约束项保拉普拉斯坐标项缩放约束项Q为测地距离权重矩阵，为所求解的变形图，为待求解的变形图顶点集合，为的顶点个数，为基于耦合准调和基进行姿态迁移的结果，为旋转变换矩阵，δi G为变形图G的拉普拉斯坐标，为所求解变形图的拉普拉斯坐标，为的缩放矩阵，λ1、λ2为标量权值。4. a kind of spectral attitude migration method based on deformation map according to claim 1, is characterized in that: in step 3), after utilizing coupling quasi-harmonic basis to carry out attitude migration, the attitude migration result is utilized to optimize energy function Find the deformation graph after pose migration, where the vertex constraint term Paula Place coordinate term scaling constraints Q is the geodesic distance weight matrix, is the solved deformation diagram, is the set of deformed graph vertices to be solved, for the number of vertices, is the result of pose transfer based on coupled quasi-harmonic basis, is the rotation transformation matrix, δ i G is the Laplace coordinate of the deformed graph G, is the Laplace coordinate of the solved deformation graph, for The scaling matrix of , λ 1 and λ 2 are scalar weights. 5.根据权利要求1所述的一种基于变形图的谱姿态迁移方法，其特征在于：在步骤4)中，将求解得到的变形图利用嵌入变形图编辑方法变形源网格生成目标网格，包括以下步骤：5. a kind of spectral pose migration method based on deformation map according to claim 1, is characterized in that: in step 4) in, the deformation map that the solution obtains utilizes embedded deformation map editing method to deform source grid to generate target grid , including the following steps: 4.1)利用最小化能量函数优化姿态迁移结果得到源网格变形的旋转平移矩阵，其中为的顶点个数，旋转项单个顶点旋转项Rot(Rj)＝(c1·c2)2+(c1·c3)2+(c2·c3)2+(c1·c2)2+(c1·c1-1)2+(c2·c2-1)2+(c3·c3-1)2,c1、c2和c3是变形图顶点处旋转矩阵Rj的列向量，规则化项N(j)为顶点vj的邻域，t为平移变换，为变形图顶点，αjn＝1.0，顶点约束项 为生成的目标网格上的顶点，为变形图的顶点，wrot、wreg、wcon为标量权值；4.1) Utilize the minimized energy function Optimize the pose transfer result to obtain the rotation-translation matrix of the deformation of the source mesh, where for the number of vertices, the rotation term Single vertex rotation term Rot(R j )=(c 1 ·c 2 ) 2 +(c 1 ·c 3 ) 2 +(c 2 ·c 3 ) 2 +(c 1 ·c 2 ) 2 +(c 1 · c 1 -1) 2 +(c 2 ·c 2 -1) 2 +(c 3 ·c 3 -1) 2 , c 1 , c 2 and c 3 are the column vectors of the rotation matrix R j at the vertex of the deformed graph, regularization term N(j) is the neighborhood of vertex v j , t is the translation transformation, is the deformation graph vertex, α jn = 1.0, the vertex constraint term for the generated target mesh apex on deformation map The vertices of , w rot , w reg , and w con are scalar weights; 4.2)得到旋转平移矩阵后利用公式计算目标网格的顶点集其中wj(vl')为测地距离权重，为旋转矩阵，为平移矩阵，vl'为源网格的顶点，gj(vi')是与源网格顶点vi'测地距离最近的m个变形图G上的顶点集合，为所求最终目标网格上的顶点。4.2) After obtaining the rotation and translation matrix, use the formula Calculate the target grid vertex set of where w j (v l ') is the geodesic distance weight, is the rotation matrix, is the translation matrix, v l ' is the vertex of the source mesh, g j (v i ') is the set of vertices on m deformation graphs G with the closest geodesic distance to the vertex v i ' of the source mesh, is the vertex on the desired final target mesh. 6.根据权利要求1所述的一种基于变形图的谱姿态迁移方法，其特征在于：在步骤5)中，分割参考网格、生成的目标网格及其变形图姿态迁移不充分区域，根据步骤2)至步骤4)进行分层姿态迁移，直到姿态迁移充分为止，包括如下步骤：6. a kind of spectral pose migration method based on deformation map according to claim 1, is characterized in that: in step 5) in, segmentation reference grid, generated target grid and deformation map posture migration insufficient area, Carry out hierarchical pose migration according to steps 2) to 4) until the pose migration is sufficient, including the following steps: 5.1)选取姿态学习不充分的局部网格,将对应的参考网格M、生成的目标网格M′、变形图进行分割,得到对应的局部子网格S、S′和Gs；5.1) Select the local grid with insufficient attitude learning, and combine the corresponding reference grid M, the generated target grid M′, and the deformation map. Segmentation is performed to obtain the corresponding local subgrids S, S ′ and Gs; 5.2)利用步骤2)至步骤3)求出局部变形图Gs姿态迁移后的变形图 5.2) Use steps 2) to 3) to obtain the deformation map of the local deformation map G s after the pose transfer 5.3)计算Gs分割边界顶点位置在姿态迁移前后的平均偏移量，将其添加到的相应顶点；5.3) Calculate the average offset of G s segmentation boundary vertex positions before and after pose transfer, add it to the corresponding vertex of ; 5.4)利用的顶点信息更新变形图 5.4) Utilize Update the deformation graph with the vertex information of 5.5)利用步骤4)得到最终姿态迁移结果。5.5) Use step 4) to obtain the final pose transfer result.
CN201910056771.5A 2019-01-22 2019-01-22 Spectral attitude migration method based on deformation graph Active CN109859322B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910056771.5A CN109859322B (en) 2019-01-22 2019-01-22 Spectral attitude migration method based on deformation graph

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910056771.5A CN109859322B (en) 2019-01-22 2019-01-22 Spectral attitude migration method based on deformation graph

Publications (2)

Publication Number Publication Date
CN109859322A true CN109859322A (en) 2019-06-07
CN109859322B CN109859322B (en) 2022-12-06

Family

ID=66895490

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910056771.5A Active CN109859322B (en) 2019-01-22 2019-01-22 Spectral attitude migration method based on deformation graph

Country Status (1)

Country Link
CN (1) CN109859322B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111027438A (en) * 2019-12-03 2020-04-17 Oppo广东移动通信有限公司 Human body posture migration method, mobile terminal and computer storage medium
CN112530016A (en) * 2020-10-30 2021-03-19 北京字跳网络技术有限公司 Method, device, equipment and storage medium for adsorbing road fittings
CN112819961A (en) * 2021-02-18 2021-05-18 桂林电子科技大学 Simplified grid deformation method and device based on micro-computation

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6362820B1 (en) * 1999-06-24 2002-03-26 Microsoft Corporation Quadric metric for simplifying meshes with appearance attributes
CN1842824A (en) * 2004-08-03 2006-10-04 松下电器产业株式会社 Person determination device and person search and track device
CN101071514A (en) * 2006-05-12 2007-11-14 中国科学院自动化研究所 Method for directly transferring three-dimensional model attitude
CN102402691A (en) * 2010-09-08 2012-04-04 中国科学院自动化研究所 Method for tracking human face posture and motion
CN102855666A (en) * 2012-08-21 2013-01-02 北京师范大学 Craniofacial reconstructing method based on hierarchical regression model
CN103996221A (en) * 2014-04-21 2014-08-20 北京农业信息技术研究中心 Plant organ mesh simplification method targeted for visualization calculation
US20150379769A1 (en) * 2012-02-20 2015-12-31 Thomson Licensing Method and apparatus for mesh simplification
CN106484511A (en) * 2016-09-30 2017-03-08 华南理工大学 A kind of spectrum attitude moving method
CN107256557A (en) * 2017-05-03 2017-10-17 华南理工大学 A kind of controllable subdivision curved surface image vector method of error
CN107527384A (en) * 2017-07-14 2017-12-29 中山大学 A kind of lattice simplified method of Three-Dimensional Dynamic based on motion feature and its system
CN108961411A (en) * 2018-07-02 2018-12-07 南京大学 A kind of simplified method of the complex three-dimensional building model keeping external appearance characteristic

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6362820B1 (en) * 1999-06-24 2002-03-26 Microsoft Corporation Quadric metric for simplifying meshes with appearance attributes
CN1842824A (en) * 2004-08-03 2006-10-04 松下电器产业株式会社 Person determination device and person search and track device
CN101344923A (en) * 2004-08-03 2009-01-14 松下电器产业株式会社 People Search Tracking Device
CN101071514A (en) * 2006-05-12 2007-11-14 中国科学院自动化研究所 Method for directly transferring three-dimensional model attitude
CN102402691A (en) * 2010-09-08 2012-04-04 中国科学院自动化研究所 Method for tracking human face posture and motion
US20150379769A1 (en) * 2012-02-20 2015-12-31 Thomson Licensing Method and apparatus for mesh simplification
CN102855666A (en) * 2012-08-21 2013-01-02 北京师范大学 Craniofacial reconstructing method based on hierarchical regression model
CN103996221A (en) * 2014-04-21 2014-08-20 北京农业信息技术研究中心 Plant organ mesh simplification method targeted for visualization calculation
CN106484511A (en) * 2016-09-30 2017-03-08 华南理工大学 A kind of spectrum attitude moving method
CN107256557A (en) * 2017-05-03 2017-10-17 华南理工大学 A kind of controllable subdivision curved surface image vector method of error
CN107527384A (en) * 2017-07-14 2017-12-29 中山大学 A kind of lattice simplified method of Three-Dimensional Dynamic based on motion feature and its system
CN108961411A (en) * 2018-07-02 2018-12-07 南京大学 A kind of simplified method of the complex three-dimensional building model keeping external appearance characteristic

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
MENGXIAOYIN ET AL: "Spectral pose transfer", 《COMPUTER AIDED GEOMETRIC DESIGN》 *
张天序等: "三维运动目标的多尺度智能递推识别新方法", 《自动化学报》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111027438A (en) * 2019-12-03 2020-04-17 Oppo广东移动通信有限公司 Human body posture migration method, mobile terminal and computer storage medium
CN112530016A (en) * 2020-10-30 2021-03-19 北京字跳网络技术有限公司 Method, device, equipment and storage medium for adsorbing road fittings
CN112819961A (en) * 2021-02-18 2021-05-18 桂林电子科技大学 Simplified grid deformation method and device based on micro-computation
CN112819961B (en) * 2021-02-18 2023-08-22 桂林电子科技大学 Simplified Mesh Deformation Method and Device Based on Differentiable Computation

Also Published As

Publication number Publication date
CN109859322B (en) 2022-12-06

Similar Documents

Publication Publication Date Title
CN109035388A (en) Three-dimensional face model method for reconstructing and device
CN102054296A (en) Grid deformation method based on local rigidity
WO2011111680A1 (en) Method and apparatus for creating surface processing data
Xu et al. Dynamic harmonic fields for surface processing
CN109859322A (en) A kind of spectrum posture moving method based on deformation pattern
CN110751665B (en) A method and system for reconstructing 3D portrait model from portrait relief
CN109118578A (en) A kind of multiview three-dimensional reconstruction texture mapping method of stratification
CN113393577B (en) Oblique photography terrain reconstruction method
CN108230452B (en) A Model Hole Filling Method Based on Texture Synthesis
CN109118589A (en) Geometrical model full range details restorative procedure
Zheng et al. A new topology optimization framework for stiffness design of beam structures based on the transformable triangular mesh algorithm
Wang et al. Arap++: an extension of the local/global approach to mesh parameterization
CN112330798B (en) A 3D Small Object Reconstruction Method Based on Parametric Model
CN111274732B (en) Grid repairing method based on 'connection relation-position' iterative optimization
CN106484511A (en) A kind of spectrum attitude moving method
Zhou et al. Image deformation with vector-field interpolation based on MRLS-TPS
Qiu et al. An efficient and collision-free hole-filling algorithm for orthodontics
Zayer et al. Curvilinear spherical parameterization
CN113792459A (en) Mesh fairing optimization method and device for finite element triangular mesh
CN112750155A (en) Panoramic depth estimation method based on convolutional neural network
Bo et al. Surface fitting with cyclide splines
CN116958430A (en) A face modeling method, device, equipment and medium
CN111047687A (en) A solid modeling method for heterogeneous materials based on 3D T-splines
Shi et al. Self-parameterization based multi-resolution mesh convolution networks
He et al. Modeling and realization of image-based garment texture transfer

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant