A Fabric Image Segmentation Pipeline to Find the Yarn Weaving Pattern from a Single Photograph

Abstract

Rendering high quality woven fabric images need special rendering techniques that differ from standard approaches. Since a woven fabric consists of vertically and horizontally crossing elliptical yarns and yarns, consist of cylindrical microfibers, a suitable, revised rendering approach is needed. Therefore, a rendering model for woven fabric images is proposed "Microcylinder Appearance Model". The model needs some parameters to produce a realistic rendering of different fabric types such as silk, linen, velvet, and polyester satin. In the original article, authors derived these parameters via eye inspection. Our ultimate aim is to automatize this process. In this article, we propose a pipeline that does the first step of this automatization, i.e. segmenting the captured image via image processing techniques and finding the yarn crossing points. The segmented image can further be utilized to extract parameters of cylindrical shading models in future works. © 2020 ACM.