Title :
A new hierarchical simultaneous localization and mapping method based on local shape context of environment
Author :
Ronghua Luo ; Huaqing Min ; Yan Shao
Author_Institution :
Dept. of Comput. Sci. & Eng., South China Univ. of Technol., Guangzhou, China
Abstract :
A new kind of local shape context (LSC) is proposed for hierarchical simultaneous localization and mapping (HSLAM) by taking account of the features of information from laser range Ander. In the process of HSLAM, LSC is used to incrementally segment the map that has been created into topological node. And at the same time an active loop closing method for HSLAM is proposed based on the information of LSC. Experiments are carried out with the robot designed by our lab and the experimental results prove that our method performs well for map building in large environments.
Keywords :
SLAM (robots); laser ranging; mobile robots; shape recognition; active loop closing method; hierarchical simultaneous localization and mapping method; laser range finder; local shape context; robot; Buildings; Silicon; Simultaneous localization and mapping; active loop closing; hierarchical map representation; local shape context; simultaneous localization and mapping;
Conference_Titel :
Machine Learning and Cybernetics (ICMLC), 2010 International Conference on
Conference_Location :
Qingdao
Print_ISBN :
978-1-4244-6526-2
DOI :
10.1109/ICMLC.2010.5580817
