m12-DR 1.1.1: Acquisition of spatial maps of semi-structured environments

Building the representation of the environment of a robot is essential to two general tasks: autonomous navigation and communication of the perception of the robot to the end-user. This report presents the results of the work- package 1 on spatio-temporal modeling for situation awareness during the first year. The objectives were to build consistent spatial representation of USAR sites (MS1.1) based on multi-modal localization in semi-structured environments and hybrid maps. We used state-of-the-art metrical 2D map- ping algorithm on which we developed an advanced topological decomposi- tion method based on information theory. We show results on 3D perception dedicated to SLAM. We also advanced the tools for modular and efficient Iterative Closest Point (ICP) algorithm.

DR111-PUBLIC.pdf — PDF document, 33712Kb