Recognition and re‐visualization of woven fabric structures

Purpose The purpose of this paper is to develop a computerized program that can recognize woven fabric structures and simultaneously use the obtained data to 3D re‐visualize the corresponding woven fabric structures. Design/methodology/approach A 2D bitmap image of woven fabric was initially acquired using an ordinary desktop flatbed scanner. Through several image‐processing and analysis techniques as well as recognition algorithms, the weave pattern was then identified and stored in a digital format. The weave pattern data were then used to construct warp and weft yarn paths based on Peirce's geometrical model. Findings By combining relevant weave parameters, including yarn sizes, warp and weft densities, yarn colours as well as cross‐sectional shapes, a 3D image of yarns assembled together as a woven fabric structure is produced and shown on a screen through the virtual reality modelling language browser. Originality/value Woven fabric structures can now be recognised and simultaneously use the obtained data to 3D re‐visualize the corresponding woven fabric structures.