Attention-based active 3D point cloud segmentation
M. Johnson-Roberson · Jeannette Bohg · Marten Bjorkman · D. Kragic
TL;DR
A min-cut method labels multiple objects in a 3D point cloud at once, with several seeding strategies compared on real robot-head data.
BRIEF
Robots that work around everyday objects need to tell where one object ends and another begins. This work segments several objects at once from a 3D point cloud, rather than from a flat image. It uses a min-cut graph method to assign every point a label globally, extending earlier work that separated a single object from its background. The authors also compare several ways of seeding the optimization. Tests use real data from an active vision robotic head, with aggregate numbers and visual examples.
Abstract
In this paper we present a framework for the segmentation of multiple objects from a 3D point cloud. We extend traditional image segmentation techniques into a full 3D representation. The proposed technique relies on a state-of-the-art min-cut framework to perform a fully 3D global multi-class labeling in a principled manner. Thereby, we extend our previous work in which a single object was actively segmented from the background. We also examine several seeding methods to bootstrap the graphical model-based energy minimization and these methods are compared over challenging scenes. All results are generated on real-world data gathered with an active vision robotic head. We present quantitive results over aggregate sets as well as visual results on specific examples.