- publication
- Symposium on Vision, Modeling and Visualization (VMV)
- authors
- Katja Wolff, Philipp Herholz, Olga Sorkine-Hornung
Our method improves visual appearance for garments with automatically aligned textures along seams by incorporating the reflection symmetry of the human body. Our input is a sewing pattern and the desired fabric print - a 2D wallpaper pattern (a). Recent work by Wolff and Sorkine-Hornung [WS2019] describes a method to align repetitive textile prints along seams and optimizes this fit by slightly deforming the sewing pattern. The resulting texture alignment, sewing pattern and 3D garment shape are not globally symmetric (b). We expand this existing method and incorporate the global reflection symmetry of the human body, as defined by the user. The resulting sewing pattern and 3D garment shape are symmetric and show improved visual appearance (c). Even though we increase the number of constraints, the quality of the texture alignment along individual seams is not visibly affected. The mismatch at each seam edge is visualized on a color scale from green (0 cm) to red (2 cm).
abstract
Recent work in the area of digital fabrication of clothes focuses on repetitive print patterns, specifically the 17 wallpaper groups, and their alignment along garment seams. While adjusting the underlying sewing patterns for maximized fit of wallpapers along seams, past research does not account for global symmetries that underlie almost every sewing pattern due to the symmetry of the human body. We propose an interactive tool to define such symmetries and integrate them into the existing algorithm, such that both the texture alignment and the deformation of the sewing pattern adhere to these symmetries.