Instead of using an expensive 3D Laser scanner, a 3D volumetric representation of environments for the Simultaneous Localization and Mapping (SLAM) with a 2D Laser Range Finder, augmented with an additional self-integrated servomechanism was devised. Since the computing time of Iterative Closest Points (ICP) algorithm mainly depends on the numbers of cloud points, we proposed an effective edge feature extraction method to extract the edge points and reduce the redundant points in a 3D scene. Experiment results have shown its efficiency.
Published in:
Automation and Logistics (ICAL), 2010 IEEE International Conference on
Date of Conference: 16-20 Aug. 2010