CN114647969A - Cloth parameter measuring method and system - Google Patents

Abstract

The invention discloses a cloth parameter measurement method and system, comprising: acquiring a pre-deformation image and a post-deformation image of a target cloth; determining a first coordinate of a key point of the target cloth before deformation according to the pre-deformation image; the key point is an artificial mark point; The deformed image determines the second coordinate of the key point after the target cloth is deformed; the deformation gradient of the key point is determined according to the first coordinate and the second coordinate; the deformation gradient of each key point is determined according to the cloth parameters, the deformation gradient of the key point and the Green strain tensor model. The initial assignment of the cloth parameters to each key point is randomly performed; the Green strain tensor model is used to iteratively optimize the cloth parameters to obtain the optimal values of the cloth parameters; the optimal values of the cloth parameters are used for cloth simulation. The cloth parameter measurement method of the present invention uses the Green strain tensor model to model the cloth, and the obtained cloth parameters have high accuracy, thereby improving the authenticity of the modeling.

Description

Method and system for measuring cloth parameters

technical field

The invention relates to the technical field of virtual fitting, in particular to a method and system for measuring cloth parameters.

Background technique

With the rapid development of VR technology in recent years, human beings have a strong interest in virtual technology. Coupled with the rise of major online shopping platforms, virtual fitting has emerged as the times require. It allows users to experience the experience without going to a physical store. The effect of clothing on the body, and unlike most of the clothing produced in the current fashion industry, it can be personalized to modify clothing according to personal preferences, and truly achieve personalized clothing without leaving home. The virtual fitting involves the measurement of fabric parameters, which requires parametric modeling of the fabric. At present, the mass-spring model is commonly used to model the movement of the fabric, but the authenticity of its modeling is still lacking.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method and system for measuring cloth parameters.

For achieving the above object, the present invention provides the following scheme:

A method for measuring cloth parameters, comprising:

Obtain the pre-deformation image and the post-deformation image of the target cloth;

Determine the first coordinates of the key point before the deformation of the target cloth according to the image before deformation; the key point is an artificial mark point;

Determine the second coordinate of the deformed key point of the target cloth according to the deformed image;

Determine the deformation gradient of the key point according to the first coordinate and the second coordinate;

Randomly perform initial assignment of cloth parameters to each of the key points;

Determine the force of each key point according to the cloth parameter, the deformation gradient of the key point and the Green strain tensor model;

Determine the loss function value according to the force of each key point;

Adjust the cloth parameters according to the loss function value, and jump to the step of determining the force of each key point according to the cloth parameters, the deformation gradient of the key points and the Green strain tensor model, until the loss function value When the iteration stop condition is met, the jump is stopped, and the optimal value of the cloth parameter is obtained; the optimal value of the cloth parameter is used for cloth simulation.

Optionally, determining the force of each key point according to the cloth parameter, the deformation gradient of the key point and the Green strain tensor model, specifically includes:

Determine the Green strain tensor of each of the key points in the deformed image of the target cloth according to the deformation gradient of the key point;

Determine the warp and weft direction of the target fabric;

determining the stiffness matrix of the target cloth according to the cloth parameters;

Determine the Green's stress tensor of each of the key points in the deformed image of the target cloth according to the Green's strain tensor and the stiffness matrix;

Determine the strain energy density corresponding to each of the key points in the deformed image of the target cloth according to the Green strain tensor and the Green stress tensor;

Determine the strain energy of the target cloth after deformation according to the strain energy density;

Obtain the acceleration of each of the key points according to the strain energy;

determine the quality of each of said key points;

The force of each key point is calculated according to the acceleration and the mass of the key point.

Optionally, using Adam adaptive gradient descent method to iteratively optimize the cloth parameters.

Optionally, the target cloth is a mesh cloth, and the key point is a junction point of the grid; the determining the deformed strain energy of the target cloth according to the strain energy density specifically includes: placing the target The image grid before the cloth is deformed or the image grid after the target cloth is deformed is divided into triangular grids along the diagonal;

Determine the strain energy of the triangular mesh according to the strain energy density and the area of the triangular mesh where the key point is located;

The deformed strain energy of the target cloth is determined according to the strain energy of each of the triangular meshes.

Optionally, a least square method is used to obtain the coordinates of each of the key points.

A cloth parameter measurement system, the system includes:

The image acquisition module is used to acquire the pre-deformation image and the post-deformation image of the target cloth;

a first coordinate acquisition module, configured to determine the first coordinate of the key point before the deformation of the target cloth according to the image before deformation; the key point is an artificial mark point;

A second coordinate acquisition module, configured to determine the second coordinates of the deformed key points of the target cloth according to the deformed image;

a deformation gradient determination module, configured to determine the deformation gradient of the key point according to the first coordinate and the second coordinate;

The initial assignment module of the cloth parameters is used to randomly assign the initial assignment of the cloth parameters to each of the key points;

The force determination module of the key points is used to determine the force of each of the key points according to the cloth parameters, the deformation gradient of the key points and the Green strain tensor model;

The loss function value determination module is used to determine the loss function value according to the force of each key point;

The module for determining the optimal value of the cloth parameter is used to adjust the cloth parameter according to the loss function value, and jump to the method of determining each key point according to the cloth parameter, the deformation gradient of the key point and the Green strain tensor model. In the stressing step, when the loss function value satisfies the iterative stop condition, the jump is stopped, and the optimal value of the cloth parameter is obtained; the optimal value of the cloth parameter is used for the cloth simulation.

Optionally, the force determination module of the key point specifically includes: determining the Green strain tensor of each key point in the deformed image of the target cloth according to the deformation gradient of the key point; Warp and weft direction; determine the stiffness matrix of the target fabric according to the fabric parameters; determine the Green stress tensor of each of the key points in the deformed image of the target fabric according to the Green strain tensor and the stiffness matrix; Determine the strain energy density corresponding to each of the key points in the deformed image of the target cloth according to the Green strain tensor and the Green stress tensor; determine the strain energy of the target cloth after deformation according to the strain energy density ; obtain the acceleration of each key point according to the strain energy; determine the mass of each key point; calculate the force of each key point according to the acceleration and the mass of the key point.

Optionally, using Adam adaptive gradient descent method to iteratively optimize the cloth parameters.

Optionally, the target cloth is a mesh cloth, and the key point is a junction point of the grid; the determining the deformed strain energy of the target cloth according to the strain energy density specifically includes: placing the target The image grid before the cloth is deformed or the image grid after the target cloth is deformed is divided into triangular grids along the diagonal;

Determine the strain energy of the triangular mesh according to the strain energy density and the area of the triangular mesh where the key point is located;

The deformed strain energy of the target cloth is determined according to the strain energy of each of the triangular meshes.

Optionally, a least square method is used to obtain the coordinates of each of the key points.

According to the specific embodiments provided by the present invention, the following technical effects are disclosed: obtaining the pre-deformation image and the post-deformation image of the target cloth; determining the first coordinates of the pre-deformation key points of the target cloth according to the pre-deformation images; the key points are artificial mark point; determine the second coordinate of the deformed key point of the target cloth according to the deformed image; determine the deformation gradient of the key point according to the first coordinate and the second coordinate; initial assignment of the cloth parameters; determine the force of each key point according to the cloth parameter, the deformation gradient of the key point and the Green strain tensor model; determine the loss function value according to the force of each key point; according to the The loss function value adjusts the cloth parameter, and jumps to the step of determining the force of each key point according to the cloth parameter, the deformation gradient of the key point and the Green strain tensor model, until the loss function value satisfies the iteration stop When conditions are met, the jump is stopped, and the optimal value of the cloth parameter is obtained; the optimal value of the cloth parameter is used for cloth simulation. The cloth parameter measurement method of the present invention uses the Green strain tensor model to model the cloth, and the obtained cloth parameters have high accuracy, thereby improving the authenticity of the modeling.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings required in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some of the present invention. In the embodiments, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative labor.

1 is a schematic flowchart of a method for measuring a cloth parameter provided in Embodiment 1 of the present invention;

2 is a picture after the target cloth provided in Embodiment 1 of the present invention is deformed by an external force exerted by a weight;

3 is a schematic diagram of triangular mesh division provided in Embodiment 1 of the present invention;

Fig. 4 is the action diagram of the test function v provided in Embodiment 1 of the present invention;

5 is an effect diagram obtained by using the Green strain tensor model optimization provided in Embodiment 1 of the present invention;

Fig. 6 is the effect diagram obtained by adopting the mass spring model optimization provided by Embodiment 1 of the present invention;

7 is an effect diagram of simulation using the Green strain tensor model and the optimal value of the second cloth parameter provided in Embodiment 1 of the present invention;

FIG. 8 is an effect diagram of simulation provided by the embodiment 1 of the present invention using the mass spring model and the cloth parameters obtained by optimization in FIG. 6;

9 is a simulation effect diagram of hanging cloth by using cloth parameters optimized by the Green strain tensor model of the present invention;

10 is a simulation effect diagram of cloth falling on a sphere using cloth parameters optimized by the Green strain tensor model of the present invention;

FIG. 11 is a structural diagram of a cloth parameter measurement system provided in Embodiment 2 of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The purpose of the present invention is to provide a cloth parameter measurement method and system. The method adopts the Green strain tensor model to model the cloth, and the obtained cloth parameters have high accuracy, thereby improving the authenticity of the modeling.

In order to make the above objects, features and advantages of the present invention more clearly understood, the present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

Example 1

The present invention provides a method for measuring cloth parameters, referring to FIG. 1 , the method includes the following steps:

Step 101: Acquire a pre-deformation image and a post-deformation image of the target cloth.

In this embodiment, the target cloth is cut into a square along a certain warp and weft direction, and the cloth is deformed by applying external force to the cloth by using fishing line and weight, and the image of the target cloth before the deformation and the target cloth are obtained by shooting with a binocular camera. The deformed image, the deformed image is shown in Figure 2, and the target cloth can be cloth of any material.

Step 102: Determine the first coordinates of the key point before the deformation of the target cloth according to the image before deformation; the key point is an artificial mark point.

Step 103: Determine the second coordinates of the deformed key points of the target cloth according to the deformed image.

In this embodiment, Matlab is used for corner detection to obtain the pixel coordinates of the intersection points of the cloth grids in the two photos. These intersection points are also called key points, and through these key points, the piece of cloth can be reconstructed. In order to convert the pixel coordinates of the junction point to world coordinates, this requires the use of internal and external parameters of the camera, as shown in the following formula:

Among them, the two matrices on the right side of the equal sign are the internal parameter matrix and the external parameter matrix of the camera, respectively, f _x and f _y in the internal parameter matrix are the focal length of the camera, x ₀ and y ₀ are the main point coordinates, and s is the coordinate axis tilt Parameters, the R element of 3x3 in the external parameter matrix is the rotation matrix of the camera, and the column vector of 3x1 in the external parameter matrix is the translation matrix of the camera. The internal and external parameter matrix of the camera can be obtained by the way of camera calibration, u, v are the key points The pixel coordinates of , and X, Y, and Z are the world coordinates of the obtained key points. If the product of the internal and external parameter matrices is the m matrix, the following formula is shown:

Among them, the elements in the m matrix are the multiplication of the corresponding row and column elements of the camera's internal and external parameter matrix. The formula (1) can be written as:

After expanding the matrix in (3), we can get:

(um ₃₁ -m ₁₁ )X+(um ₃₂ -m ₁₂ )Y+(um ₃₃ -m ₁₃ )Z=m ₁₄ -um ₃₄ (4)

(vm ₃₁ -m ₂₁ )X+(vm ₃₂ -m ₂₂ )Y+(vm ₃₃ -m ₂₃ )Z=m ₂₄ -vm ₃₄ (5)

For this keypoint, there are currently two equations, but three unknowns to solve. Since two cameras are used, two of the above equations, four equations are obtained, which becomes the solution of the overdetermined equations. The world coordinates of the key point can be obtained by using the least squares method.

Cut the fabric along different warp and weft directions to obtain multiple square fabrics, obtain multiple sets of pre-deformed and post-deformed images of the target fabric through the binocular camera, and repeat steps 101-103 to obtain the coordinates of several sets of key points . These groups are randomly divided into training group and test group, the training group is used to optimize the cloth parameters, and the test group is used to test the quality of the optimized cloth parameters.

Step 104: Determine the deformation gradient of the key point according to the first coordinate and the second coordinate.

Step 105: Randomly perform initial assignment of cloth parameters to each of the key points.

Step 106: Determine the force of each of the key points according to the cloth parameters, the deformation gradient of the key points, and the Green strain tensor model.

Step 107: Determine the loss function value according to the force of each key point.

Step 108: Adjust the cloth parameters according to the loss function value, and jump to the step of determining the force of each key point according to the cloth parameters, the deformation gradient of the key points, and the Green strain tensor model, until the When the loss function value satisfies the iterative stop condition, the jump is stopped, and the optimal value of the cloth parameter is obtained; the optimal value of the cloth parameter is used for cloth simulation.

In this embodiment, steps 104 to 108 are the process of determining the optimal value of the cloth parameters, and the details are as follows:

Use the world coordinates of the key points of the training group obtained in step 103 before and after deformation as input to model the cloth, and the modeling method is as follows:

Connect the input key points into a square grid, and then connect the diagonal lines of the square grid as shown in Figure 3 to divide the cloth into many triangular grids, such as △ABC and △BCD, which maintain the basic shape of the cloth , controls the stretching of the cloth. Although the existence of these triangular meshes can resist the bending of the cloth perpendicular to the BC edge, they cannot resist the bending of the cloth along the BC edge, so another type of triangular mesh needs to be introduced. Connect another diagonal of the square mesh to get another set of triangle meshes in different directions, such as △ABD and △ACD, which can resist the bending of the cloth along the BC edge, and of course they also control the stretching of the cloth. The cooperation of these two types of triangular meshes can control the stretching and bending of the cloth, which will make the deformed cloth tend to change towards the shape before the deformation. Triangular meshes are generated for the key points before deformation and the key points after deformation, a total of two groups. For each triangular mesh before deformation, take the center of gravity of the triangular mesh as the origin, and use the triangular mesh plane as the XOY plane to establish a two-dimensional material coordinate system, so as to calculate the material coordinates of the three key points of the triangular mesh Since the material coordinate system is two-dimensional, the calculated material coordinate system coordinates are also two-dimensional. For each deformed triangular mesh, the three-dimensional world coordinate system coordinates of the key points obtained in step 103 are used.

Using Green's strain tensor to calculate the force on key points, the formula is as follows:

(dL) ² -(dl) ² =dxdx-dudu=Fdu*Fdu-dudu=du ^T (F ^T FI)du=2du ^T *∈*du (6)

where dL and dl are the lengths of the line elements after deformation and before deformation, respectively, while dx and du are the line elements after deformation and before deformation, respectively, they are a pair of vectors, and F is the deformation gradient, I is the unit matrix, and ∈ is the key Green's strain tensor of points. Further, the deformation gradient is equal to the derivative of the coordinates of the world coordinate system after the deformation of the line element to the coordinates of the material coordinate system before the deformation, and the following formula can be obtained:

where x is the coordinate of the world coordinate system after the line element is deformed, u is the coordinate of the material coordinate system before the line element is deformed, and δ is the Kronecker symbol. In the Green strain tensor model, the material coordinate system coordinates before the deformation of the three key points of the triangular grid obtained above are used, and the barycentric coordinates of the triangular grid can represent the material coordinate system coordinates of any point in the triangular grid before the deformation. , the formula is as follows:

Among them, u is the material coordinate system coordinate before the deformation of the arbitrary point, m is the material coordinate system coordinate before the deformation of the three key points of the triangular mesh, and b is the barycentric coordinate. The coordinates of the world coordinate system after the deformation of the arbitrary point can also be expressed as follows in the same way:

Wherein x is the coordinates of the world coordinate system after the deformation of the arbitrary point, and p is the coordinates of the world coordinate system after the deformation of the three key points of the triangular mesh, that is, the second coordinates in step 103 . Simultaneous equations (8) and (9) yield:

in:

The deformation gradient of this arbitrary point can be expressed as follows:

Where: δ _i =[δ _i1 δ _i2 0] (14)

When the δ subscripts are the same, the value of δ is 1, otherwise the value of δ is 0.

It can be seen from the above formula that the deformation gradient of any point is independent of the position of the arbitrary point, that is, the deformation gradient of any point in the same triangular mesh is the same as the deformation gradient of the key point.

According to the Green's strain tensor and stiffness matrix of the key points obtained above, determine the Green's stress tensor of each of the key points in the deformed image of the target cloth, and the calculation process is as follows:

Among them, ∈ ₁₁ and ∈ ₂₂ are line strains, and ∈ ₁₂ are shear strains, which are obtained from the above calculations, and σ ₁₁ , σ ₂₂ and σ ₁₂ are the corresponding three Green stresses. The parameters in the stiffness matrix are the cloth parameters that need to be measured, and there are nine in total. Cloth has the characteristics of anisotropy, and the stiffness matrix corresponding to anisotropic materials is generally an asymmetric matrix, but considering that this will increase the complexity of the entire cloth model, from the perspective of engineering practicality, this is often ignored. asymmetry, and treat it as a symmetric matrix, the symmetric stiffness matrix is as follows:

The stiffness matrix is further simplified, and the simplified stiffness matrix is as follows:

Among them, c _ij is the influence factor of the deformation in the j direction on the force generated in the i direction.

In this embodiment, the initial assignment of the cloth parameters is randomly performed to each of the key points, specifically: the measured warp and weft directions are six, which are divided according to the angle between the warp of the cloth and the coordinate axis of the world coordinate system, and the order is 0 °, 30°, 60°, 90°, 120° and 150°, 96 initial values of cloth parameters are randomly generated. First determine the warp direction of the entire target cloth, then calculate the angle between the warp direction and the coordinate axis of the world coordinate system, and finally use the angle to perform linear interpolation in the cloth parameters of the six initially generated warp and weft directions to obtain the target Parameters in the cloth stiffness matrix. Since the parameters in the stiffness matrix are properties of the cloth, all triangle meshes should share a stiffness matrix.

According to the Green strain tensor and Green stress tensor obtained above, determine the strain energy density corresponding to each of the key points in the deformed image of the target cloth, and the formula is as follows:

Among them, ψ is the strain energy density corresponding to the key point, ∈ is the Green's strain tensor of the key point, and σ is the Green's stress tensor of the key point.

Determine the strain energy of the target cloth after deformation according to the strain energy density. The Green's stress tensor and Green's strain tensor at any point in a triangular mesh are the same, so the calculated strain energy density is also the same, so the integral will be The strain energy of the triangular grid is obtained by multiplying the strain energy density by the area of the triangular grid. Then, the strain energies of all triangular meshes can be added to obtain the strain energy of the target cloth after deformation.

The force calculation is performed using the Lagrangian mechanics equation. The complete Lagrangian mechanics equation is as follows:

where x is the coordinate of the world coordinate system after the deformation of the line element, u is the coordinate of the material coordinate system before the deformation of the line element,

为形变后的线元在世界坐标系中的速度，Q为线元受到的外力，L为所述目标布料的拉格朗日量，其表达式如下：

is the velocity of the deformed line element in the world coordinate system, Q is the external force received by the line element, L is the Lagrangian of the target cloth, and its expression is as follows:

L=T-V (20)

的函数。考虑到L对于整块布料空间而言并非是连续的，至少从三角网格交界处来看，它并不是连续的，所以引入弱形式来对它进行求解，式子如下：Wherein, T is the kinetic energy of the target cloth, and V is the potential energy of the target cloth. x is transformed by u, so x is a function of u; if the coordinates of the material coordinate system before the deformation of the line element are determined, the external force on the line element can be obtained, so Q is a function of u; and the Lagrangian quantity represents the energy of the cloth after deformation, so L is x and

The function. Considering that L is not continuous for the entire cloth space, at least from the point of view of the triangular mesh junction, it is not continuous, so a weak form is introduced to solve it, the formula is as follows:

The weak form weakens equality everywhere. In order to equalize the sum of the integrated areas in the target cloth, a test function v is introduced to enhance this formula to express the area that can be weakened to the integral in a narrow region where the value of the test function v is non-zero. The sum is equal:

The role of the test function v is shown in Figure 4. Write L in the form of T-V, since the derivative of kinetic energy to position is zero, and the derivative of potential energy to velocity is zero, so substitution can be obtained:

The above formula is the derivative of the functional, which is defined by the formula of the derivative of the functional:

Expressions with kinetic and potential energies:

V=∫ _w (Ψ+ρg ^T x)du (26)

where ρ is the areal density of the cloth, Ψ is the strain energy output in step 4, and g is the acceleration of gravity. Finally, we can get:

表示形变后的世界坐标系坐标对形变前的材料坐标系坐标x分量的偏导数，

表示形变后的世界坐标系坐标依次对形变前的材料坐标系坐标x和y分量的偏导数，以此类推。然后引入有限元法来将连续的目标布料离散为三角网格。对于布料空间中任意点的坐标，可用关键点的坐标来表示，如下：in

Represents the partial derivative of the coordinates of the world coordinate system after deformation to the x component of the coordinates of the material coordinate system before deformation,

Indicates the partial derivatives of the coordinates of the world coordinate system after deformation to the x and y components of the coordinates of the material coordinate system before deformation, and so on. Then the finite element method is introduced to discretize the continuous target cloth into triangular meshes. The coordinates of any point in cloth space can be represented by the coordinates of key points, as follows:

表示第I个关键点形变后的世界坐标系坐标，N_I表示第I个关键点的权重，之所以用关键点形变前的材料坐标系坐标作为自变量，是因为不管布料空间中的顶点如何发生形变，该关键点坐标的权重不会改变。在之前的弱形式中，引入了测试函数v，并说明它是一个时而为零时而非零的函数，用于将积分分为一个个狭小的区域，以便增强弱形式。这里的权重就具有这样的性质，当点位于该关键点组成的三角网格中时，权重非零，其余情况权重均为零，所以另v＝N_K，可以得到如下式子：in

Represents the world coordinate system coordinates of the I-th key point after deformation, and N _I represents the weight of the I-th key point. The reason why the material coordinate system coordinates before the key point deformation is used as an independent variable is because regardless of the vertices in the cloth space. Deformation occurs, and the weight of the key point coordinates does not change. In the previous weak form, the test function v was introduced and stated to be a function that is sometimes zero but not zero to divide the integral into small regions in order to strengthen the weak form. The weight here has such a property. When the point is located in the triangular mesh composed of the key point, the weight is non-zero, and in other cases, the weight is zero, so the other v=N _K , the following formula can be obtained:

M _KI =∫ _w ρN _I N _K du (30)

where _MKI is the quality matrix. Bringing in the strain energy of the target cloth obtained above after deformation, the acceleration of each key point after deformation can be calculated.

In this embodiment, the effective area of a key point is determined by the triangular mesh that contains it. Since the quality of the key points should not change with the deformation of the triangular mesh, the triangular mesh before deformation is used for the calculation. Calculate the initial area of each triangular mesh containing the key point, divide each area by three to obtain the effective area of each triangular mesh belonging to the key point, and then add all the effective areas to be the key The effective area of the point is finally multiplied by the density of the cloth material to obtain the quality of each of the key points.

Then, the force of each key point is calculated according to the acceleration of the key point and the mass of the key point obtained above.

In this embodiment, the process of calculating the loss function value in steps 107 and 108 and adjusting the cloth parameters according to the loss function value can be as follows:

The loss value corresponding to the cloth parameter is calculated as follows:

Where k _n is the parameter obtained by the nth iteration optimization, I is the number of key points, and f _i is the resultant force received by the i-th key point after deformation. The loss function calculates the average value of the resultant force of all key points. The smaller the loss function, the more quasi-static the cloth is. After that, the Adam adaptive gradient descent method and the PSO particle swarm algorithm are used to optimize the loss function. The specific optimization process is as follows:

Step S1: First, the Adam adaptive gradient descent method is used for optimization iteration. The initial step size is selected as 0.01, and the compensation is reduced by ten times every 10,000 iterations. The remaining parameters are set to 0.9, 0.999 and 1e-8 according to the standard. Take the 96 parameters randomly generated in step 105 as the initial values, repeat steps 104-107, that is, calculate the strain energy of the deformed cloth, the force of the key points after the deformation, and the loss function value to optimize these parameters to obtain the first cloth parameter The optimal value.

Step S2: The optimal value of the first cloth parameter obtained in step S1 is further optimized by using the PSO particle swarm algorithm. The specific parameters of the PSO algorithm are 50 particles, and the initial speed of the search is a random number between -0.01 and 0.01, and finally the optimal value of the second cloth parameter will be obtained.

Figure 5 shows the result obtained by using the optimal value of the second cloth parameter, and the loss is 0.0016. For comparison, the mass-spring model is optimized with the same number of parameters. Figure 6 shows the final result obtained by optimizing the mass-spring model, with a loss of 0.0211. Obviously, the Green strain tensor model is more accurate than the mass spring model for cloth modeling with the same number of parameters.

The following uses the test group to verify the cloth parameters obtained in step 108:

Step 1: Use the world coordinates of the key points of the test group for simulation. There is no internal force in the first simulation iteration, that is, the strain energy of the cloth obtained in step 106 is 0, because the cloth does not deform, but because the cloth is affected by external force, Get the acceleration of non-zero keypoints. Simulation iterations are performed using Verlet explicit integration, which is based on the Taylor expansion, as follows:

x(t+δt)=2x(t)-x(t-δt)+a(t)δt ² +O(δt ⁴ ) (34)

x(t+δt)=2x(t)-x(t-δt)+a(t)δt ² (35)

Ignoring high-order terms, after determining the time step, the position of the next simulation iteration of the key point is obtained through the position and acceleration of the current simulation iteration of the key point, and the time step is 0.1ms.

Step 2: Set a certain number of simulation iterations, or check whether the displacement of most key points in two adjacent simulation iterations is less than a given threshold to stop the simulation iteration, and finally get the simulated positions of all key points.

Step 3: Calculate the difference between the simulated positions of all the key points obtained in step 2 and the world coordinates of the deformed key points in the test group as a second norm, and use the calculation result as the standard to measure the optimal value of the second cloth parameter. . As shown in Figure 7, the effect of the simulation using the Green strain tensor model and the optimal value of the second cloth parameter, the average deviation displacement of the key points obtained by the simulation iteration relative to the test group groundtruth is 4.36mm. Figure 8 is the effect diagram of the simulation using the mass-spring model and the parameters optimized in Figure 6. The average deviation displacement of the key points obtained by the simulation iteration relative to the test group groundtruth is 7.08mm. Therefore, with the same number of parameters, for For cloth modeling, the Green strain tensor model is more accurate than the mass-spring model.

Fig. 9 and Fig. 10 simulate the cloth based on the cloth parameters obtained by the present invention, Fig. 9 is the simulation effect diagram of the hanging cloth, and Fig. 10 is the simulation effect diagram of the cloth falling on the sphere.

Example 2

The present invention also provides a cloth parameter measurement system, see FIG. 11 , the system includes:

The image acquisition module 201 is used for acquiring the pre-deformation image and the post-deformation image of the target cloth.

The first coordinate acquisition module 202 is configured to determine the first coordinate of the key point of the target cloth before deformation according to the image before deformation; the key point is an artificial marker point.

The second coordinate acquisition module 203 is configured to determine the second coordinates of the deformed key points of the target cloth according to the deformed image.

A deformation gradient determination module 204, configured to determine the deformation gradient of the key point according to the first coordinate and the second coordinate.

The initial assignment module 205 of the cloth parameters is used to randomly perform the initial assignment of the cloth parameters to each of the key points.

The force determination module 206 of the key points is configured to determine the force of each of the key points according to the cloth parameters, the deformation gradient of the key points and the Green strain tensor model.

The loss function value determination module 207 is used to determine the loss function value according to the force of each key point.

A cloth parameter optimal value determination module 208, configured to adjust the cloth parameter according to the loss function value, and jump to determine each key point according to the cloth parameter, the deformation gradient of the key point and the Green strain tensor model The step of stressing, until the loss function value satisfies the iteration stop condition, stop jumping, and obtain the optimal value of the cloth parameter; the optimal value of the cloth parameter is used for cloth simulation.

In this embodiment, the force determination module of the key point specifically includes: determining the Green strain tensor of each key point in the deformed image of the target cloth according to the deformation gradient of the key point; determining the target The warp and weft direction of the cloth; the stiffness matrix of the target cloth is determined according to the cloth parameters; the Green stress tension of each key point in the deformed image of the target cloth is determined according to the Green strain tensor and the stiffness matrix determine the strain energy density corresponding to each of the key points in the deformed image of the target cloth according to the Green strain tensor and the Green stress tensor; determine the deformed target cloth according to the strain energy density strain energy; obtain the acceleration of each key point according to the strain energy; determine the mass of each key point; calculate the force of each key point according to the acceleration and the mass of the key point.

In this embodiment, the system uses the Adam adaptive gradient descent method to iteratively optimize the cloth parameters.

In this embodiment, the target cloth is a mesh cloth, and the key points are the junction points of the grid; determining the strain energy of the target cloth after deformation according to the strain energy density, specifically including: placing the target cloth The image mesh before the cloth deformation or the image mesh after the target cloth deformation is diagonally divided into triangle meshes; the strain energy of the triangle mesh is determined according to the strain energy density and the area of the triangle mesh where the key points are located; The deformed strain energy of the target cloth is determined according to the strain energy of each of the triangular meshes.

In this embodiment, the system uses the least squares method to obtain the coordinates of each of the key points.

For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant part can be referred to the description of the method.

The principles and implementations of the present invention are described herein using specific examples. The descriptions of the above embodiments are only used to help understand the method and the core idea of the present invention; meanwhile, for those skilled in the art, according to the present invention There will be changes in the specific implementation and application scope. In conclusion, the contents of this specification should not be construed as limiting the present invention.

Claims (10)

1. a cloth parameter measuring method, is characterized in that, comprises: Obtain the pre-deformation image and the post-deformation image of the target cloth; Determine the first coordinates of the key point before the deformation of the target cloth according to the image before deformation; the key point is an artificial mark point; Determine the second coordinate of the deformed key point of the target cloth according to the deformed image; Determine the deformation gradient of the key point according to the first coordinate and the second coordinate; Randomly perform initial assignment of cloth parameters to each of the key points; Determine the force of each key point according to the cloth parameter, the deformation gradient of the key point and the Green strain tensor model; Determine the loss function value according to the force of each key point; Adjust the cloth parameters according to the loss function value, and jump to the step of determining the force of each key point according to the cloth parameters, the deformation gradient of the key points and the Green strain tensor model, until the loss function value When the iteration stop condition is met, the jump is stopped, and the optimal value of the cloth parameter is obtained; the optimal value of the cloth parameter is used for cloth simulation. 2 . The method according to claim 1 , wherein the determining the force of each of the key points according to the cloth parameters, the deformation gradient of the key points and the Green strain tensor model specifically includes: 2 . Determine the Green strain tensor of each of the key points in the deformed image of the target cloth according to the deformation gradient of the key point; Determine the warp and weft direction of the target fabric; determining the stiffness matrix of the target cloth according to the cloth parameters; Determine the Green's stress tensor of each of the key points in the deformed image of the target cloth according to the Green's strain tensor and the stiffness matrix; Determine the strain energy density corresponding to each of the key points in the deformed image of the target cloth according to the Green strain tensor and the Green stress tensor; Determine the strain energy of the target cloth after deformation according to the strain energy density; Obtain the acceleration of each of the key points according to the strain energy; determine the quality of each of said key points; The force of each key point is calculated according to the acceleration and the mass of the key point. 3 . The method according to claim 1 , wherein the cloth parameters are iteratively optimized by Adam adaptive gradient descent method. 4 . 4. The method according to claim 2, wherein the target cloth is a mesh cloth, and the key point is a junction point of the mesh; The determining of the deformed strain energy of the target cloth according to the strain energy density specifically includes: dividing the image mesh of the target cloth before deformation or the mesh of the target cloth after deformation into triangular meshes along a diagonal line; Determine the strain energy of the triangular mesh according to the strain energy density and the area of the triangular mesh where the key point is located; The deformed strain energy of the target cloth is determined according to the strain energy of each of the triangular meshes. 5 . The method according to claim 1 , wherein the coordinates of each of the key points are obtained by using the least squares method. 6 . 6. A cloth parameter measurement system, characterized in that, comprising: The image acquisition module is used to acquire the pre-deformation image and the post-deformation image of the target cloth; a first coordinate acquisition module, configured to determine the first coordinate of the key point before the deformation of the target cloth according to the image before deformation; the key point is an artificial marker point; A second coordinate acquisition module, configured to determine the second coordinates of the deformed key points of the target cloth according to the deformed image; a deformation gradient determination module, configured to determine the deformation gradient of the key point according to the first coordinate and the second coordinate; The initial assignment module of the cloth parameters is used to randomly assign the initial assignment of the cloth parameters to each of the key points; The force determination module of the key points is used to determine the force of each of the key points according to the cloth parameters, the deformation gradient of the key points and the Green strain tensor model; The loss function value determination module is used to determine the loss function value according to the force of each key point; The module for determining the optimal value of the cloth parameter is used to adjust the cloth parameter according to the loss function value, and jump to the method of determining each key point according to the cloth parameter, the deformation gradient of the key point and the Green strain tensor model. In the stressing step, when the loss function value satisfies the iterative stop condition, the jump is stopped, and the optimal value of the cloth parameter is obtained; the optimal value of the cloth parameter is used for the cloth simulation. 7 . The system according to claim 6 , wherein the force determination module of the key points specifically comprises: determining, according to the deformation gradient of the key points, the value of each of the key points in the deformed image of the target cloth. 8 . Green strain tensor; determine the warp and weft direction of the target cloth; determine the stiffness matrix of the target cloth according to the cloth parameters; determine the deformed image of the target cloth according to the Green strain tensor and the stiffness matrix the Green's stress tensor of each of the key points; the strain energy density corresponding to each of the key points in the deformed image of the target cloth is determined according to the Green's strain tensor and the Green's stress tensor; according to the strain energy Density determines the strain energy of the target cloth after deformation; obtains the acceleration of each key point according to the strain energy; determines the mass of each key point; calculates each key point according to the acceleration and the mass of the key point stress at key points. 8. The system according to claim 6, wherein the cloth parameters are iteratively optimized using Adam adaptive gradient descent method. 9 . The system according to claim 7 , wherein the target cloth is a mesh cloth, and the key point is a junction point of the mesh; 9 . The determining of the deformed strain energy of the target cloth according to the strain energy density specifically includes: Dividing the image grid of the target cloth before deformation or the image grid of the target cloth after deformation into triangular grids along the diagonal; Determine the strain energy of the triangular mesh according to the strain energy density and the area of the triangular mesh where the key point is located; The deformed strain energy of the target cloth is determined according to the strain energy of each of the triangular meshes. 10 . The system according to claim 6 , wherein the coordinates of each of the key points are obtained by using a least square method. 11 .

Images