主催: 一般社団法人 日本機械学会
会議名: ロボティクス・メカトロニクス 講演会2023
開催日: 2023/06/28 - 2023/07/01
The accurate 3-D map construction applicable to a robot navigation on outdoor irregular terrains is promising technology for future disaster response. We have performed the experiments of the 3-D mapping of such irregular terrains equipped in Fukushima Robot Test Field using the man-powered movable wagon. The 3-D maps are constructed by LIDAR-SLAM coupled with the IMU data, and color information are provided to the maps by data fusion with the multiple camera images. Two types of the irregular terrains of about 100 -200 m scales are almost exactly restored as 3-D colored point cloud map, although larger one of the global maps is not closed well. Improvement of the mapping accuracy will be expected by correction of the robot self-location using additional positioning data such as RTK-GNSS.