A hierarchical approach to high-quality partial shape registration

This paper presents a feature-driven, hierarchical shape registration algorithm. The central idea is to generate correspondences in multiple levels in a coarse-to-fine manner, with additional features incrementally inserted in each level. The registration starts from the coarsest resolution. Registration results obtained in one level serve as references for the registration in the next level. We adopt the heat kernel coordinates [3] for local shape parameterization, giving rise to a complete solution capable of registering partial shapes undergoing isometric deformation with higher accuracy. Through experiments, we demonstrate the effectiveness of this multi-resolution method and its advantages over the single-resolution method.