“Optimization of cloth simulation parameters by considering static and dynamic features” by Kunitomo, Morishima and Nakamura

Realistic drape and motion of virtual clothing is now possible by using an up-to-date cloth simulator, but it is even difficult and time consuming to adjust and tune many parameters to achieve an authentic looking of a real particular fabric. Bhat et al. [2003] proposed a way to estimate the parameters from the video data of real fabrics. However, this projects structured light patterns on the fabrics, so it might not be possible to estimate the accurate value of the parameters if fabrics have colors and textures. In addition to the structured light patterns, they use a motion capture system to track how the fabrics move. In this paper, we will introduce a new method using only a motion capture system by attaching a few markers on fabric surface without any other devices. Moreover, animators can easily estimate the parameters of many kinds of fabrics with this method. Authentic looking and motion of simulated fabrics are realized by minimizing error function between captured motion data and synthetic motion considering both static and dynamic cloth features.

K. S. Bhat, C. D. Twigg, J. K. Hodgins, P. K. Khosla, Z. Popovic and S. M. Seitz. “Estimating Cloth Simulation Parameters from Video”, SCA, pp37–51, 2003. Google ScholarDigital Library