CLAM, Collaborative Localization and Mapping Using Fiducial System and Exchangeable Range Sensors
Author: Jiayi Liu
Publisher:
Published: 2012
Total Pages: 246
ISBN-13:
DOWNLOAD EBOOKTraditional vision based localization and mapping frameworks rely heavily on the visual features in the environment. This thesis research proposes a Collaborative Localization and Mapping (CLAM) framework that eliminates the environmental dependency of robot localization and mapping. By using the visual fiducial system and a "leap-frogging" style robot motion, localization is achieved by accumulating inter-robot displacements provided by the fiducial system. This method also enables the functionality of exchangeable range sensing. Preliminary studies are conducted to characterize the real world performance of the fiducial system in order to improve the localization accuracy. Experimental results obtained in various distinctive indoor environments are shown and discussed. The proposed approach allows robot agents to localize in ill-posed environments and also provides a way of graceful degradation for robots dysfunctional in localization.range s