Vehicle-mounted twin laser scan system for simultaneous localization and mapping (SLAM) over uneven terrain

• Main Authors: Song-Cheng Ho, 何嵩晟
• Other Authors: Herve Capart
• Format: Others
• Language: en_US
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/03871331602678467789

碩士 === 國立臺灣大學 === 土木工程學研究所 === 93 === The present thesis is devoted to the development of an imaging system for an underwater tracked vehicle. The objective is to be able to map the topography of river or coastal floors, and at the same time to position the vehicle with respect to this topography. This is known in robotics as the Simultaneous Localization and Mapping (SLAM) problem. For underwater applications, special challenges arise due to the irregular nature of the topography and the low visibility. This effectively limits the imaging scope to the immediate neighbourhood of the vehicle. To address these challenges, the system proposed is based on two parallel laser sheets imaged by a single camera and scanning the ground in sequence as the vehicle progresses. Three chapters outline respectively the conceptual design, mathematical modeling, and experimental testing of this novel approach. The experimental tests are conducted outdoors, using an instrumented carriage to scan the topography of a textured bridge surface.