Determination and quantitative description of hollow body in point cloud

Anonymous

Keywords: point cloud, voxel, connectivity, hollow body

Abstract: When volume 3d display system deals with point cloud data with hollow bodies, the hollow body area cannot be determined correctly causing the lack of color information in responding area. Existing researches lack a solution to determine hollow bodies. This paper firstly gives a quantitative description of hollow body and defines a set of parameters to describe size,shape,position of hollow bodies. Then this paper proposes a voxel connectivity regional-growth hollow body determination algorithm(VCRHD) to determine the hollow bodies in 3D point cloud. The algorithm has two phases. The first phase is to use a small amount of voxels to realizes the voxelization of the point cloud and calculate the approximate volume ratio of hollow bodies based on the voxel connectivity regional-growth principle. Then this paper uses the approximate volume ratio to determine the optimal number of voxels based on the experimental result. The second phase is to use the optimal number of voxels to determine the hollow bodies and calculate hollow body parameters which is proved to be efficient and accurate. In addition,this paper establishes a data set containing 287 different point cloud files in 7 different categories to test algorithm. The experimental results prove the feasibility of the algorithm. Finally,this paper analyzes the limitations of the algorithm and looks forward to the application prospect in the future.

TL;DR: This paper proposes the concept of hollow body and an algorithm to determine hollow bodies in point cloud.