Research On Simulation For Three Dimensional Textile Texture Deformation

Simulation for 3D cloth textile deformation has been an ever-increasing hot topic in computer graphic field. Textile has many attributes including flexibility, drapability, its material feature, and the vision angle of an observer. And many other factors should be taken into account when cloth textile simulation is performed, such as different material attributes, its mechanic attributes and the realism. The cloth textile simulation has been much more complex due to these factors. In recent years, much work has been done upon cloth textile modeling and simulating depending on the different methods. In order to improve the better display effect instead of decreasing its computation, most methods are on the base of 3D cloth modeling and simulating. Due to the computation speed, there has not a universal algorithm. The simulation for 3D cloth texture has been much more complex. The application of 3D cloth texture deformation simulation covers CAM, Virtual Fashion Shop, Virtual Reality, 3D games, Digital Video. With the development of the increasing demand to satisfy the customer, it seems important to develop more and more new methods, novel technologies.This paper proposes, on the base of what the previous work has been done, several efficient methods to analyze and realize the simulation for cloth textile texture deformation. And this paper also presents the 3D cloth texture deformation modeling techniques. The main contributions are introduced as follows.First of all, this study proposes a 3-D cloth texture deformation modeling and simulating based upon a wind noise method. Start from a 3-D cloth model, we analysis its texture geometrical model configuration, and take a wind noise method as the texture deformation simulation method according to the flexibility and the vision effect of the selected model. Considered that the final results should be improved the stochastic attributes, we then use and adopt a 1/f noise as the basis geometrical description formulation. Besides, the system is interpolated some control functions with human interferences. The random 3-D model may be arrived by adjusting and controlling the corresponding parameters. In order to improve the cloth texture deformation realism, we adopt an improved Phong illumination model by adjusting its parameters so that it may be suitable for the cloth texture geometrical attributes. Some experiments show that such method is effective.Secondly,this study proposes a 3-D cloth texture deformation model and simulation method based upon a neural network dynamic meshes generation. Due to the meshes numbers and shapes of the finite element method are constant and cannot improve to meet the new situations, the errors will exits when we simulate the cloth texture deformation. To find a problem-solving method, we try our best to find out a neural network model to realize the meshes regeneration upon the finite element meshes: the proper meshes refining operation should be adopted and employed where the large deformations occur in order to represent the texture minor deformation attributes; the proper meshes re-combination operation should be adopted where the cloth deformation seems small or disappear so that the meshes numbers may be reduced and the effect may not be affected. Here,the mesh configuration we adopt is called quadrilateral structure. Combining with the advantages of the neural network and the finite element, such method is what we try as we can to find out to realize 3-D cloth texture deformation simulation and the results may also be able to meet users'meets.Thirdly, this study addresses a cloth texture deformation simulation technique based on a texture topological structure representation. The cloth texture may be represented by its color,position, and topological structure. Its texture color relationship may be calculated before deformation occurs. Then, we turn to discuss the plane texture deformation. The texture 3-D deformation representation may be therefore obtained. On the one hand, we should represent the cloth texture planar deformation. To solve such problem, we may adopt a curve-fitting method to fit the texture surface according to the actual deformation. Depends upon the distance of the coordinate origin passing along the curve and the texture topological structure, we may calculate its texture color. On the other hand, in order to calculate the 3-D cloth texture colors, we attempt to find out the differences between the controlled curves' distances before and after deformation on the bicubic b-spline surface. Furthermore, the vision angle parameters are also under considered.Such method may be proved by taking an example of applying a neural network meshes generation and dynamic meshes generation methods into it.Fourthly, this study addresses cloth wrinkle 3-D attributes extraction based upon an image analysis and recognition. The detailed description may be as follows: 1) a cloth wrinkle 3-D attributes extraction method is proposed by the different intensities and the height value information of the different regions of the initial cloth wrinkle image. Furthermore, the key points on cloth wrinkled surface should be found out and positioned. These points have the highest intensities. According to these points' information, we may calculate the incident angle of light source. On the controlled curved surface, each point's light intensity and height value may be calculated; 2) this paper proposes a method to extract cloth wrinkle 3-D attributes on the basis of a human-interfere improved genetic algorithm. By comparing the extracted cloth 3-D wrinkle attribute with that in the original image, the differences should be found out and therefore the next iteration development can then be determined. In genetic algorithm improvement, the relationship between parents and offspring may be coded as genes, and by these genes can we realize our method to extract cloth wrinkle's boundary, height and shaded areas. Under the external forces, the cloth begins to deform and to be in motion. The genetic algorithm's adaptive function may be configured depending upon each geometrical model extracted. The relationship between the captured angle and each image can be found out eventually.This study made some research on three dimensional cloth texture deformation simulations.The concepts mentioned in this paper contribute to some cloth texture deformation display simulation methods. Furthermore, it may put the development of the technology forward in such field. Therefore, it has somewhat theoretical research meaning. Besides, it may also propose some theoretical supports and applicable techniques for the computer graphic research, intelligent calculation and aerodynamics model in the fabric technology research field. It also makes some contributions to the development of computer graphic and virtual reality, and hereby it is important and valuable to be realized of the virtual fashion development and corresponding digital movie makers, the computer games designers, virtual reality software, and so on.