Raising Local Density for Object Reconstruction using Auxiliary Points

This investigation proposes a novel method for identifying high-risk point sets before surface reconstruction. Given a set of points P sampled from some unknown surface M, triangular meshes are constructed to approximate the scanned model. The point sets are obtained by laser scanner sampling on the object, so increasing the variance of the object shape increases the likelihood of irregular density distribution and the occurrence of non-indigenous holes. To solve this problem, this study proposes an algorithm to raise density, and to judge whether a new sampling point is required. The add-point algorithm discovers the locations of add-points, and maps them to the MLS surface on the basis of "feature points extracted" and "non-feature points with insufficient density" point sets. The test results demonstrate that the proposed strategy of adding auxiliary points improves the point distribution, and helps prevent the formation of non-indigenous holes. Finally, the auxiliary points are removed and returned to the original point set