SPaM: soft patch matching for non-rigid pointcloud registration

Behnam Maleki, Raphael Falque, Teresa A. Vidal-Calleja, Alen Alempijevic

Abstract: We propose a novel non-rigid registration framework for raw, unstructured, deformable point clouds purely based on geometric features. Our method does not rely on visual information nor a pre-defined shape template. The representation of the scene consists simply of a point cloud. In this approach, the global non-rigid deformation of an object is formulated as an aggregation of locally rigid transformations. The concept of locality is embodied in soft patches described by geometrical properties based on SHOT descriptor and its neighborhood. By considering the confidence score of pairwise association between soft patches of two scans (not necessarily consecutive), the computed similarity matrix serves as the seed to grow a correspondences graph. We take advantage of the rigidity terms defined in As-Rigid-As-Possible for pruning and optimizing this correspondences graph. Individual points are then transformed proportionally to their distance with respect to their adjacent patch centroids, given the optimized graph and the weighted average of the associated transformations. Experiments demonstrate the capability of the proposed approach to cope with large deformations blended with numerous missing parts in the scan process.