Ground plane estimation from sparse LIDAR data for loader crane sensor fusion system

Abstract

Research on the development of control systems for loader cranes, despite their importance in the industry, is conducted by only a few scientific centers. West Pomeranian University of Technology, Szczecin in collaboration with loader cranes manufacturer — Cargotec company, started research on the multisensory monitoring system for cranes. proposed system also allows you to track the position of the operator. This paper presents the subsystem for ground plane estimation and ground points filtration. The developed algorithm uses data from the Velo dyne LIDAR VLP-16 scanner. The subsystem is designed for real time operation. It is based on the RANSAC algorithm and vector dot product. The effectiveness of the algorithm was compared with other algorithms described in this publication. Tests were carried out on a loader crane test bench at various positions of the LIDAR sensor. Experiments confirms that ground plane estimation results of the proposed algorithm are better than other presented methods.