An Effective Feature-Preserving Mesh Simplification Scheme Based on Face Constriction

JH Wu,SM Hu,CL Tai,JG Sun
DOI: https://doi.org/10.1109/pccga.2001.962853
2001-01-01
Abstract:A novel mesh simplification scheme that uses the face constriction process is presented. By introducing a statistical measure that can distinguish triangles having vertices of high local roughness from triangles in flat regions into our weight-ordering equation, along with other heuristics, our scheme can better preserve visually important features in the original mesh. To improve the shape quality of triangles, we adopt nonlinear face area sensitivity in the weight ordering. A learning and feedback mechanism is also utilized to enhance user controllability. The computations are simple, making our scheme time-effective and easy to implement. In addition to comparing our scheme with other mesh simplification algorithms empirically, we compare their performances by establishing a unifying ground among three basic simplification processes: decimate vertex, collapse edge, and constrict face. This unification allows us to analyze the intrinsic merits and demerits of simplification algorithms to help users make better selections.
What problem does this paper attempt to address?