Multi-resolution geometric fusion

Abstract

Geometric fusion of multiple sets of overlapping surface measurements is an important problem for complete 3D object or environment modelling. Fusion based on a discrete implicit surface representation enables fast reconstruction for complex object modelling. However, surfaces are represented at a single resolution resulting in impractical storage costs for accurate reconstruction of large objects. This paper addresses accurate reconstruction of surface models independent of object size. An incremental algorithm is presented for implicit surface representation of an arbitrary triangulated mesh in a volumetric envelope around the surface. A hierarchical volumetric structure is introduced for efficient representation by local approximation of the surface within a fixed error bound using the maximum voxel size. Multi-resolution geometric fusion is achieved by incrementally constructing a hierarchical surface representation with bounded error. Results are presented for validation of the multi-resolution representation accuracy and reconstruction of real objects. Multi-resolution geometric fusion achieves a significant reduction in representation cost for the same level of geometric accuracy.