Hole Boundary Detection of a Surface of 3D Point Clouds

Processing surfaces with data coming from an automatic acquisition technique always generates the problem of unorganized 3D point sets or presence of geometric deficiencies (i.e. in some regions of the surface, the obtained data points are sparse or without containing any points). It leads to what we call "holes", and a mandatory surface reconstruction process. Applying the process to the whole set of 3D points in order to explore and detect all these holes that can be served to process in a further step. In this paper, we propose a method to detect the hole-boundary on a surface of 3D point clouds structured in a 3D volume. The method consists of two steps. In the first step, we extract the exterior boundary of the surface based on the neighborhood relationship between the 3D points and our particular definition of these points. In the second step, we detect the hole boundary by using a growth function on the interior boundary of the surface. Our method could process very fast and extract exactly the boundary of all holes on the surface.