When observing or controlling a robot remotely over high latency connections, it is important to compress and limit the amount of data being sent without sacrificing the amount of information being provided to the client. For many robots, tf topics are quite large and are published frequently. wviz_tf_manager publishes a topic tf_changes that publishes the tfs that have changed at a given polling frequency. A tf is considered to have changed if the change is larger than either the translational_update_distance or the angular_update_distance (parameters that are bot set by default to 10). By default the polling_frequency is set to 10 hz but can be set as a rosparam in the launch file. Additionally the node provides a service, publish_all_transforms, that allows a client to request all transforms.

Running the wviz_tf_manager node

  • $ rosrun wviz_tf_manager wviz_tf_manager


Published Topics

tf_changes (tf/tfMessage)
  • tf messages which have recently been updated


~<name>/publish_all_transforms (wviz_tf_manager/PublishAllTransforms)
  • Allows a client to request that all available tfs be published


~<name>/polling_frequency (double, default: 10)
  • Frequency at which changed tfs are published
~<name>/translational_update_distance (double, default: 10)
  • Minimum amount of translational change required
~<name>/angular_update_distance (double, default: 10)
  • Minimum amount of angular change required

Wiki: wviz_tf_manager (last edited 2011-04-01 22:06:58 by SarahOsentoski)