A Novel Algorithm for Periodic Conformal Flattening of Genus-one and Multiply Connected Genus-zero Surfaces
Zhong-Heng Tan, Tiexiang Li, Wen-Wei Lin, Shing-Tung Yau
TL;DR
The paper introduces periodic conformal flattening to map genus-one and multiply connected genus-zero surfaces onto fundamental periodic domains by minimizing a discrete conformal energy, producing a sparse quadratic problem. It yields two algorithms: DPCF for genus-one and SPCF for multiply connected surfaces, both solving via small linear systems and independent of cut-path choices. The authors prove cut-path independence and bijectivity under intrinsic Delaunay conditions, with an IDT preprocessing strategy to guarantee bijectivity when needed. Numerical experiments show 4–5× efficiency gains over prior methods and robust texture-mapping results, illustrating practical impact in mesh processing and graphics. The work highlights limitations to higher-genus generalization and points to ongoing extension toward broader applicability.
Abstract
In this paper, we propose a novel parameterization method for genus-one and multiply connected genus-zero surfaces, called periodic conformal flattening. The conformal energy minimization technique is utilized to compute the desired conformal map, which is characterised as an easily solvable quadratic functional minimization problem, yielding a sparse linear system. The advantages of the proposed algorithms DPCF and SPCF are a) independence from the cut path selection, which introduces no additional conformal distortion near the cut seams; b) bijectivity guaranteeing for intrinsic Delaunay triangulations. The numerical experiments illustrate that DPCF and SPCF express high accuracy and a 4-5 times improvement in terms of efficiency compared with state-of-the-art algorithms.Based on the theoretical proof of the bijectivity guaranteeing, a simple strategy is applied for to guarantee the bijectivity of the resulting maps for non-Delaunay triangulations. The application on texture mapping illustrates the practicality of our developed algorithms.
