In this thesis, a “graphics for vision” approach is proposed to tackle the problem of surface reconstruction from a large and imperfect data set: surface reconstruction on demand by tensor voting (ROD-TV). ROD-TV simultaneously delivers good efficiency and robustness by adapting to a continuum of primitive connectivity, view dependence, and levels of detail (LOD). Locally inferred surface elements are robust to noise and better capture local shapes. By positioning and inferring per-vertex normals at sub-voxel precision on the fly, we can achieve interpolative shading to produce superior quality rendering. Since this missing information can be recovered at the present levels-of-detail, our result is not upper bounded by the scanning resolution....[ Read more ]

In this thesis, a “graphics for vision” approach is proposed to tackle the problem of surface reconstruction from a large and imperfect data set: surface reconstruction on demand by tensor voting (ROD-TV). ROD-TV simultaneously delivers good efficiency and robustness by adapting to a continuum of primitive connectivity, view dependence, and levels of detail (LOD). Locally inferred surface elements are robust to noise and better capture local shapes. By positioning and inferring per-vertex normals at sub-voxel precision on the fly, we can achieve interpolative shading to produce superior quality rendering. Since this missing information can be recovered at the present levels-of-detail, our result is not upper bounded by the scanning resolution.

ROD-TV consists of a spatial hierarchical data structure that encodes various levels of detail. The local surface reconstruction algorithm is tensor voting. It is applied on demand to the visible subset of data at a desired levels-of-detail, by traversing the data hierarchy and collecting tensorial support in a neighborhood. We compare our approach and present encouraging results.