The "msg.covariance" field represents the uncertainty in the deterministic field "msg.pose".
This uncertainty is described by a normal distribution centered on the pose field so the probability of the state x follows: p(x)~ N(msg.pose,msg.covariance)
Of course there is others ways to represent the pose uncertainty. For instance the PoseArray msg (particles) makes possible describing a more complex probabilistic distribution. In any case it has some drawbacks, most of algorithms based on particles usually have less computational efficence.
Off the top of my head, I can only think of one node that might use it. It's interpreted by different nodes in different fashions. The covariance represents the trustworthiness of a measurement. Many programs don't care about this value, and each one that does handles it differently.
Asked: 2011-12-14 05:58:07 -0500
Seen: 1,066 times
Last updated: Feb 15 '12
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
