Stitch meshing

Abstract

We introduce the first fully automatic pipeline to convert arbitrary 3D shapes into knit models. Our pipeline is based on a global parametrization remeshing pipeline to produce an isotropic quad-dominant mesh aligned with a 2-RoSy field. The knitting directions over the surface are determined using a set of custom topological operations and a two-step global optimization that minimizes the number of irregularities. The resulting mesh is converted into a valid stitch mesh that represents the knit model. The yarn curves are generated from the stitch mesh and the final yarn geometry is computed using a yarn-level relaxation process. Thus, we produce topologically valid models that can be used with a yarn-level simulation. We validate our algorithm by automatically generating knit models from complex 3D shapes and processing over a hundred models with various shapes without any user input or parameter tuning. We also demonstrate applications of our approach for custom knit model generation using fabrication via 3D printing.

TilinGNN

“…The first two are (i) a random search framework that uses Algorithm 1 but replaces the TilinGNN output probabilities with random values in [0, 1]; and (ii) a greedy strategy that follows existing shape packing methods (e.g., [Kwan et al 2016]) to iteratively select the tile that shares the longest boundary with the current partial tiling solution. Besides, we employed two stateof-the-art integer programming solvers: (iii) a specialized integer programming solver, Gurobi [2020], which has been shown to perform fast in many combinatorial optimization problems [Luo et al 2015;Peng et al 2019;Wu et al 2018]; and (iv) a branch & cut mixedinteger solver (Coin-BC) [2018]. To employ them, we formulate our problem using a constraint modeling language, Minizinc [csp 1999], in which we model the selection of each node as a binary variable, set the tile overlaps as hard constraints, and define an objective function (similar to L a and L e in Section 5.3 but without logarithms) to maximize the tiling coverage and total length of shared edge segments.…”

Knittable Stitch Meshes

“…Afterward, the knitting instructions generated from the final model can be used by any knitter to fabricate the model. Therefore, combining our framework with an automated stitch mesh generation process for a given arbitrary 3D model would be an important direction for future research (Wu et al 2018).…”

“…More recently, Popescu et al (2017) presented a method that automatically generates knitting patterns for nondevelopable surfaces, and Narayanan et al (2018) and Wu et al (2018) introduced methods for converting 3D models into knit structures. The method of Narayanan et al (2018) relies on a user-defined flow field to produce machine-knittable models.…”

“…This approach allows fabricating complex 3D shapes via machine knitting, but it lacks a design interface. The method of Wu et al (2018) automatically generates stitch meshes from a given input shape, but the resulting models are not knittable.…”

A 6-Letter ‘DNA’ for Baskets with Handles

“…Fabrics that are not isonemal are not really fabrics at all, as a set of warp and weft threads (that are themselves mutually interwoven) can lift off from the rest. Contemporary computer graphic research is interested in both modeling the appearance of fabrics and designing stitch meshes to cover a three-dimensional modeled surface [8][9][10][11][12][13]. There is also an active mathematical interest in ornamental Celtic knots which cover a meshed surface [14][15][16].…”