Trends in 3D Point Cloud Contents Sampling in Mobile AR/VR Platforms

Hankyul Baek, Haemin Lee, Ji-Yeon Kim, Soyi Jung, Minseok Choi, Joongheon Kim

Abstract: In recent years, point clouds have attracted increasing attention in a variety of industries, such as autonomous driving and augmented reality. The point clouds offer a denser and more specific representation than the traditional meshes. However, processing the large point cloud contents for rendering and computation is challenging as the data size of point cloud contents is much larger than the traditional mesh contents. Hence, sampling algorithm can be a major role in these industries by reducing both computation time and storage. In this paper, a comprehensive overview of the state-of-the-art point cloud sampling methods is proposed. Different concepts of sampling approaches are identified and described in detail including algorithm based sampling, geometry based sampling and finally, mesh based sampling.