| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
I'm looking for particle filter implementation in ROS to use in mobile robot localization, but it seems the only available package is amcl (Adaptive Monte Carlo), I'm not sure is it possible to use it as particle filter or not, and if it's feasible, how?
Note: The robot (wheeled robot) provides odometry data and another data source is Kinect, that provides visual odometry data using fovis.
One way to do it is with bfl (the Bayesian Filtering Library)which is part of the Orocos project and bundled with ROS. This is what packages like robot_pose_ekf use at their core. They even provide examples of how to write a particle filter for a mobile robot in their getting started guide.
I have used this library directly when pre-existing nodes don't quite provide what I want with very little trouble. Hope that helps!
Asked: 2013-11-17 02:48:17 -0500
Seen: 3,647 times
Last updated: Nov 17 '13
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.
