[robotics-worldwide] code release: vision based navigation for micro helicopters + asctec_mav_framework update

Weiss Stephan stephan.weiss at mavt.ethz.ch
Mon May 14 12:14:40 PDT 2012

Dear Robotics Community,

we are pleased to announce our code release containing two ROS stacks for
vision based navigation for micro helicopters (and other platforms). These
ROS compatible Stacks were successfully used in the European project sFly
(http://www.sfly.org <http://www.sfly.org/>) to estimate a 6DoF pose for
long-term vision based navigation of a micro helicopter in large outdoor
environments. We will also have a few talks at ICRA on these topics, see

http://www.ros.org/wiki/ethzasl_ptam: Modified version of the well known
monocular SLAM framework PTAM (http://ewokrampage.wordpress.com/).  It has
constant computational complexity and works at 20 Hz on an ATOM 1.6 GHz
single core computer. Furthermore, it is more robust in self-similar
scenes (e.g. grass, asphalt) than the original version.

http://www.ros.org/wiki/ethzasl_sensor_fusion: Sensor fusion framework
based on an Extended Kalman Filter (EKF) for vehicle pose estimation
including intra- and inter-sensor calibration. The framework is
particularly designed to work on an Micro Aerial Vehicle (MAV) carrying an
IMU and one single camera performing visual odometry as only navigation

http://www.ros.org/wiki/asctec_mav_framework, which we are using for
communication, state estimation and position control on our AscTec
helicopters has been updated to perform state prediction for the
sensor_fusion package and to be compatible with the 2012 firmware.

Corresponding ICRA talks:

- Versatile Distributed Pose Estimation and Sensor Self-Calibration for an Autonomous MAV
A talk about our sensor fusion framework using a camera and an IMU
TuA01 Regular Session, Meeting Room 1 (Mini-sota)
Tuesday 09:30-09:45, Paper TuA01.5

- Real-Time Onboard Visual-Inertial State Estimation and Self-Calibration of MAVs in Unknown Environments
A talk about our vision framework with focus on optical flow
TuC01 Regular Session, Meeting Room 1 (Mini-sota)Tuesday 15:15-15:30, Paper TuC01.4-

- Workshop on Long-Term Autonomy : Long-Term Vision Based MAV Navigation in Unknown Environments
A talk about our multi-sensor fusion approach used in the project sFly
The River Center, St. Paul, Minnesota Friday, May 18, 2012, 8:30am--5:00pm
(Workshop), 10:55 (our talk)

Best, your ASL-sFly team
Margarita Chli, Simon Lynen, Laurent Kneip
Markus Achtelik, Stephan Weiss

More information about the robotics-worldwide mailing list