How to set the dimensions of the initial pose array
The initial pose array is really much bigger than it needs to be. How do I change it's initial dimensions?
ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
The initial pose array is really much bigger than it needs to be. How do I change it's initial dimensions?
From looking at the code, AMCL sizes the pose array in the "particlecloud" message to be the size of the number of guesses it currently has for the robot's pose. There's no way to change its initial dimensions. If you just want the current pose of the robot displayed in rviz, you can enable TF frames and look at the "base_link" frame relative to the "map" frame. The "particlecloud" message, provides a visualization that helps give insight into the particle filter AMCL uses internally.
The dimensions of the initial particle cloud should be dependent on the covariance of the initial pose estimate. Rviz won't let you change that value, but you should be able to publish it manually and try out different values for the covariance matrix until you get the initial spread you want. For example code of publishing the initial pose estimate programmatically, se https://github.com/cwru-robotics/cwru-ros-pkg/blob/master/cwru_tour_guide/tour_guide_executive/src/initial_pose_setup.py .
Asked: 2011-04-16 21:37:34 -0600
Seen: 1,729 times
Last updated: Apr 18 '11
Default 2D pose estimate when navigating
Should i be able to set the current pose for amcl?
Retrieving pose of an object from an image or bag file
How to recieve an array over Publisher and Subscriber? (Python)
Shoudl AMCL attempt to respect/get hints from walls?
Which 6D pose estimation package works best and will be supported?
How to programatically set the 2D pose on a map ?
How to handle the kinect camera's vertical movement while using amcl?