Triangulation-based plane extraction for 3D point clouds

Abstract

The processing of point clouds for extracting semantic knowledge plays a crucial role in state of the art mobile robot applications. In this work, we examine plane extraction methods that do not rely on additional point features such as normals, but rather on random triangulation in order to allow for a fast segmentation. When it comes to an implementation in this context, typically the following question arises: RANSAC or Hough transform? In this paper, we examine both methods and propose a novel plane extraction approach based on the randomized 3D Hough transform. Our main concerns for improvement are extraction time, accuracy, robustness as well as memory consumption. © Springer-Verlag Berlin Heidelberg 2012.