Deformation Sensitive Decimation

In computer graphics, many automatic methods for simplifying polygonal meshes have been developed. These techniques work very well for static meshes; however, they do not handle the case of deforming objects. Simplifying any single pose of a deforming object may destroy detail required to represent the shape in a different pose. We present an automatic technique for simplifying deforming character meshes that takes a set of examples as input and produces a set of simplified examples as output. The method preserves detail required for deformation and maintains connectivity information across the examples. This technique is applicable to many current skinning algorithms including example-driven techniques, fitting techniques and Linear Blend Skinning.