Hierarchical Structure Recovery of Point-Sampled Surfaces

We focus on the class of regular models defined by Várady et al. for reverse engineering purposes. Given a 3D surfaceM represented through a dense set of points, we present a novel algorithm that convertsM to a hierarchical representationHM. InHM, the surface is encoded through patches of various shape and size, which form a hierarchical atlas. IfM belongs to the class of regular models, thenHM captures the most significant features ofM at all the levels of detail. In this case, we show that HM can be exploited to interactively select regions of interest onM and intuitively re-design the model. Furthermore, HM intrinsically encodes a hierarchy of useful segmentations ofM. We present a simple though efficient approach to extract and optimise such segmentations, and we show how they can be used to approximate the input point sets through idealised manifold meshes.