Name
Abstract | ||
|---|---|---|
Laser scanners have been proven to provide reliable and highly precise environment perception for micro aerial vehicles (MAV). This oftentimes makes them the first choice for tasks like obstacle avoidance, close inspection of structures, self-localization, and mapping. However, artificial environments may pose problems if the scene is self-similar or symmetric and, hence, localization becomes ambiguous if only relying on distance measurements (e.g., when flying along a parallel aisle). In this paper, we propose to tackle these instances by introducing visual fiducial markers into the scene, detecting them with copter-mounted cameras and fusing these detections with laser-based self-localization in a graph optimization. Our approach abstracts the underlying multiple stages of laser-based SLAM to a slim interface that is only connected to the map building process and augments the self-localization in uncertain situations. We demonstrate the applicability of our approach during experiments in an indoor scenario with sparsely distributed fiducial markers. The test encompasses accurate map building with both the laser scanner and video cameras and subsequent relocalization relying on the detection of fiducial markers only. |
Year | DOI | Venue |
|---|---|---|
2016 | 10.1109/MFI.2016.7849554 | 2016 IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems (MFI) |
Keywords | Field | DocType |
3D laser,laser scanners,environment perception,micro aerial vehicles,MAV,visual fiducial markers,copter-mounted cameras,laser-based self-localization,graph optimization,laser-based SLAM,map building process,video cameras | Obstacle avoidance,Aisle,Computer vision,Fiducial marker,Graph optimization,Laser scanning,Visualization,Computer science,Laser,Artificial intelligence,Simultaneous localization and mapping | Conference |
ISBN | Citations | PageRank |
978-1-4673-9709-4 | 1 | 0.35 |
References | Authors | |
8 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Sebastian Houben | 1 | 40 | 7.08 |
| David Droeschel | 2 | 292 | 21.76 |
| Sven Behnke | 3 | 1672 | 181.84 |