Giters

MAPIRlab / rf2o_laser_odometry

Estimation of 2D odometry based on planar laser scans. Useful for mobile robots with innacurate base odometry. For full description of the algorithm, please refer to: Planar Odometry from a Radial Laser Scanner. A Range Flow-based Approach. ICRA 2016 Available at: http://mapir.isa.uma.es/mapirwebsite/index.php/mapir-downloads/papers/217

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Covariance not added to publisher

jsonvillanueva opened this issue · comments

commented

Around line #1041, the publisher node odom begins setting the messages for twist and pose, but the cov_odo calculation is not included. Additionally, the 9 covariance values seems to fluctuate between +- nan when played from a bag file. Is there a reason this isn't added to the publisher? This becomes problematic when attempting to use laser odometry in addition to other forms of odometry in an EKF.

commented

I am also getting this problem. I am confused why the covariance matrix is always 0. I am also wanting to use this odometer topic together with the robot_pose_ekf, but it doesn't allow for the covariance matrix to be zero. Any way to fix this?

I also saw that lots of the values are 0.13e-4 so it could be that they are not zero, just really small and they are trimmed to appear as zero. How can one overcome this problem?

commented

I asked a question related to this here:

https://github.com/MAPIRlab/mapir-ros-pkgs/issues/1