Research On Three-dimensional Reconstruction And Simulation For Knitted Fabirc

Cloth simulation is an important topic of computer graphics research. It has a wide range of applications in fashion design, television, Computer game and e-commerce. Knitting cloth as part of fabrics, on the one hand is the application needs more and more, on the other hand is facing research started late and few and can’t meet the application requirements. Whether three-dimensional display in e-commerce or fashion design, is three-dimensional simulation of cloth based on the existing fabric image. Knitted fabric is in which a larger proportion, having few research.A set of algorithms for knitted fabric form two-dimensional image reconstruction to physical simulation is proposed in our paper: first, extract information from the two-dimensional image and do knitted structure identification, second, establish knitted geometry model using NURBS curves, and optimize the physical model of knitted fabric by simplifying the force model.Finally propose an adaptive linear collision detection algorithm, which greatly improves real-time .This method has a complete set of knitted fabric from 2D to 3D.Results show that our method can ensure the reconstruction of three-dimensional model of realistic knitted fabrics, can simulate movement structural changes in the details, and meet the real-time requirements for medium-sized cloth.