ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

Revision history [back]

The problem is that there is a map->odom in the datasets we uploaded, from the online mapping phase. If postprocessing, the map->odom transform gets estimated by hector_mapping again, resulting in artifacts as tf tries to interpolate between the old and new transforms. To get rid of this problem you can filter the old transform out of the bagfiles like so:

rosbag filter Team_Hector_MappingBox_RoboCup_2011_Rescue_Arena.bag  Team_Hector_MappingBox_RoboCup_2011_Rescue_Arena_filtered.bag 'topic != "/tf" or "map" not in m.transforms[0].header.frame_id'

After doing that and playing back the filtered bagfile as described in the 'Mapping using logged data' tutorial, laser scans should be visualized correctly aligned if decay time is set to a value >0.