Active Path Planning for 3D Reconstruction with UAVs


Loading...

Author / Producer

Date

2019

Publication Type

Master Thesis

ETH Bibliography

yes

Citations

Altmetric

Data

Abstract

Autonomous mobile robots that are able to produce high quality 3D models of objects or scenes of interest are of considerable merit to a rich variety of applications such as inspection, mapping or search and rescue. In this work, a novel, realistic simulation framework as well as a package of tools for sampling based receding horizon path planning research are developed. An algorithm for fully autonomous 3D exploration and reconstruction of a target object in unknown and challenging environments under inclusion of uncertainty is presented. Using directed sensors, the impact of viewpoint orientation and distribution through space is addressed. A number of reconstruction gain and cost formulations are investigated in simulation, where a newly developed 2 stage gain formulation and an efficiency inspired cost perform best. The presented planner is thoroughly tested in simulation and quantitatively shown to outperform a state of the art method and match a manually planned path. The performance of the method is demonstrated qualitatively in a series of experiments on a real platform.

Publication status

published

External links

Editor

Contributors

Examiner: Pantic, Michael
Examiner : Khanna, Raghav
Examiner : Siegwart, Roland

Book title

Journal / series

Volume

Pages / Article No.

Publisher

ETH Zurich

Event

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Organisational unit

03737 - Siegwart, Roland Y. (emeritus) / Siegwart, Roland Y. (emeritus)

Notes

Funding

Related publications and datasets