Visual odometry, especially using a forward-looking camera only, can be challenging: It is doomed to fail from time to time and will inevitably drift in the long run. We accept this fact and present methods to cope with and correct the effects for an autonomous MAV using an RGBD camera as its main sensor. We propose correcting drift and failure in visual odometry by combining its pose estimates with information about efficiently detected ground planes in the short term and running a full SLAM back-end incorporating loop closures and ground plane measurements in pose graph optimization. We show that the system presented here achieves accurate results on several instances of the TUM RGB-D benchmark dataset while being computationally efficient enough to enable autonomous flight of an MAV.
@inproceedings{scherer2015,
title = {DCTAM: Drift-corrected tracking and mapping for autonomous micro aerial vehicles},
author = {Sebastian A. Scherer and Shaowu Yang and Andreas Zell},
booktitle = {Unmanned Aircraft Systems (ICUAS), 2015 International Conference on},
year = {2015},
address = {Denver, CO, USA},
month = {June},
pages = {1094-1101},
abstract = {Visual odometry, especially using a forward-looking camera only,
can be challenging: It is doomed to fail from time to time and will
inevitably drift in the long run. We accept this fact and present
methods to cope with and correct the effects for an autonomous MAV
using an RGBD camera as its main sensor. We propose correcting drift
and failure in visual odometry by combining its pose estimates with
information about efficiently detected ground planes in the short term
and running a full SLAM back-end incorporating loop closures and
ground plane measurements in pose graph optimization. We show that the
system presented here achieves accurate results on several instances
of the TUM RGB-D benchmark dataset while being computationally
efficient enough to enable autonomous flight of an MAV.},
days = {9-12},
pdf = {http://www.cogsys.cs.uni-tuebingen.de/publikationen/2015/scherer2015.pdf},
url = {http://dx.doi.org/10.1109/ICUAS.2015.7152401}
}