Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Mapping for Aerial Path Planning
Industrial mapping with drones requires a surface elevation model representing the heights of points in the mapped space. Such information is only sparsely available. In this project, in collaboration with Drone Harmony, we will explore ways of generating such data on the fly on a mobile device.
Keywords: 3D Reconstruction, Mobile, Mission Planning, Aerial Robots
Industrial mapping with drones in non-flat regions requires the drone to maintain a constant distance from the ground to achieve a uniform ground sample resolution (GSR). To achieve such flight in a robust way, a rough surface model representing an elevation map is needed.
While such surface models are publicly available at very low resolution globally, data of the required resolution is much harder to come by. As a result, drone operators are often required to pre-fly the region to collect imagery data, and then process it into an elevation model on a workstation. The result is then an elevation model that can be used to guide the drone's flight at an even distance from the ground.
In this project, the student will collaborate with the drone automation company Drone Harmony to develop a method to compute a simple surface model quickly on the mobile device that is controlling the drone's flight. This kind of method will allow circumventing the need to pre-fly the area and will allow the mapping mission to be completed in a single flight.
The project has a strong practical facet that will allow the student to test the designed and implemented methods in the field as part of the collaboration with Drone Harmony using commercial hardware.
Industrial mapping with drones in non-flat regions requires the drone to maintain a constant distance from the ground to achieve a uniform ground sample resolution (GSR). To achieve such flight in a robust way, a rough surface model representing an elevation map is needed.
While such surface models are publicly available at very low resolution globally, data of the required resolution is much harder to come by. As a result, drone operators are often required to pre-fly the region to collect imagery data, and then process it into an elevation model on a workstation. The result is then an elevation model that can be used to guide the drone's flight at an even distance from the ground.
In this project, the student will collaborate with the drone automation company Drone Harmony to develop a method to compute a simple surface model quickly on the mobile device that is controlling the drone's flight. This kind of method will allow circumventing the need to pre-fly the area and will allow the mapping mission to be completed in a single flight.
The project has a strong practical facet that will allow the student to test the designed and implemented methods in the field as part of the collaboration with Drone Harmony using commercial hardware.
- WP1. Learn to use the hardware and software tools needed to complete the project.
- WP2. Learn the necessary theory in the filed of 3D reconstruction.
- WP3. Design methods for the fast point cloud generation problem. This includes both operational aspects (how high to fly, whether to collect imagery or video data etc.) as well as algorithmic aspects (what method of point cloud generation to use, how to deal with sensor errors etc.).
- WP4. Implement and test designed methods in desktop and mobile environments.
- WP1. Learn to use the hardware and software tools needed to complete the project. - WP2. Learn the necessary theory in the filed of 3D reconstruction. - WP3. Design methods for the fast point cloud generation problem. This includes both operational aspects (how high to fly, whether to collect imagery or video data etc.) as well as algorithmic aspects (what method of point cloud generation to use, how to deal with sensor errors etc.). - WP4. Implement and test designed methods in desktop and mobile environments.
- Interest in drones, mapping and 3d reconstruction.
- Familiarity with Computer Vision and C++ programming.
- Interest in working with a company and perform some field work.
- Interest in drones, mapping and 3d reconstruction. - Familiarity with Computer Vision and C++ programming. - Interest in working with a company and perform some field work.
Your application must contain your most recent transcripts from bachelor and master studies. Please send the email to both supervisors. Lucas Teixeira - lteixeira@mavt.ethz.ch and David Adjiashvili (Drone Harmony): david@droneharmony.com
Your application must contain your most recent transcripts from bachelor and master studies. Please send the email to both supervisors. Lucas Teixeira - lteixeira@mavt.ethz.ch and David Adjiashvili (Drone Harmony): david@droneharmony.com