Plotting Covariance from poseWithCo​varianceSt​amped message

11 visualizaciones (últimos 30 días)
Lasse Goncz
Lasse Goncz el 4 de Mzo. de 2020
Comentada: LinuxPen el 18 de Oct. de 2022
I have recorded a bagfile and am trying to plot the /amcl_pose/pose/covariance matrix which has a message type geometry_msgs/PoseWithCovarianceStamped. I followed this tutorial on plotting data from topics, however, when i try the below code i get the error The Pose.covariance property does not exist for message type geometry_msgs/PoseWithCovarianceStamped. What am I doing wrong?
bagselect = select(bag, 'Topic', '/amcl_pose');
msgs = readMessages(bagselect);
ts = timeseries(bagselect, 'Pose.Covariance');

Respuestas (1)

Cam Salzberger
Cam Salzberger el 1 de Mayo de 2020
Hello Lasse,
I know this is a bit late to be responding to this, but maybe someone will still benefit. I believe the issue is that the use of "timeseries" with BagSelection objects is only valid for numeric, scalar properties. The "Covariance" field is an array of values, so it cannot be translated into a timeseries the way that, for example, "Pose.Pose.Point.X" would.
You should be able to see available fields by doing "timeseries" on the whole selection (or maybe limit the time first, just to make it faster) and see which data fields are available (in DataInfo.UserData).
-Cam
  1 comentario
LinuxPen
LinuxPen el 18 de Oct. de 2022
I have the same problem trying to extract the covariance matrices. @Lasse Goncz @Cam Salzberger could you please guide me on how I can plot the cov array as a function of time.
bagselect = rosbag('bagfile.bag');
bagselect3 = select(bagselect,"Topic","/amcl_pose");
cov = timeseries(bagselect3,"Pose.Covariance[0;0]","Pose.Covariance[0;1]", "Pose.Covariance[0;5]");

Iniciar sesión para comentar.

Productos

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by