Reconstrução a partir de múltiplos registros de nuvem de pontos RGB-D

Abstract

The objective of this work is to present a 3D reconstruction method using the color information. The 3D reconstruction is performed by combining point clouds obtained from di erent viewpoints. The main task is the point cloud registration algorithm that matches two point clouds. A well known algorithm for point cloud registration is the ICP (Iterative Closest Point) that determines the rotation and translation that when applied to one of the point clouds, place both point clouds in accordance. The ICP executes iteratively two main steps: point correspondence determination and registration. The point correspondence determination is a module that if not properly executed the ICP converges to a local minimum. To overcome such drawback an ICP that uses statistics to generate a dynamic distance and color threshold on the distance allowed between closest points was implemented. This approach allows subset matches, instead of matching all points from the point clouds. The surface reconstruction is performed using the marching cubes and a consensus surface algorithm with signed distance to compensate point cloud errors. In this paper the performance of the proposed method is analyzed and compared with the conventional ICP.