Weave pattern modeling of silk-like fabrics from multi-illuminated HDR image analysis

Recent advances in 3D visualization from research on computer vision (CV) and computer graphics (CG) have stimulated vigorous research on the digital modeling and rendering of woven fabrics. In this paper, we propose an efficient image-based method for modeling of weave patterns of silk-like woven fabrics of filament yarns, from multi-illuminated HDR (High Dynamic Range) image analysis, based on the mesostructure of woven fabrics, such as Noh costume. The cloth simulation results on arbitrarily chosen parameter of weave pattern model of Noh costume show the performance of the proposed approach.