Cycled merging registration of point clouds for 3D human body modeling

Abstract

In this paper, we present a cycled merging registration method based on Iterative Closest Point (ICP). We capture the point clouds by a static Kinect with the object rotating on a turntable. Different views of scan are combined by ICP and then a globally consistent human model is obtained. Our method simplifies the process of successively registration, which is usually used to solve multi-views registration from a single cycle. The main contribution of this paper is to propose a pairwise-to-global registration method, which aligns several sub-integrate views in a merging order. Our method is consistent with some cycled registration constraints which are suitable for non-rigid registration. After all point clouds are merged, the surface of the model can be estimated by Moving Least Square (MLS). A model of a part of non-rigid human body is constructed in our experiments.