Making Point Clouds with OpenDroneMap
Time: 29 January 2021 12H30 UTC (click this link to find out the time in your local time zone).
Duration: 1 hour
Venue: https://youtu.be/Zpw4XfiB47E YouTube Live Stream
Presenter: Jorge Gustavo
Host: Tim Sutton
We will use OpenDroneMap to get point clouds from images. OpenDroneMap can run from docker, so you don’t have to install anything (you must have docker, of course). We will use WebODM, which is ODM with a nice web interface.
We will cover the full life cycle, starting from surveying ground control points. Ground control points will be collected using an ArduSimple precise GPS with live corrections. It costs around 400 €. We will use SW Maps, made in Nepal. It is not open source. It is free and the author Avinab Malla answers emails. SW Maps enables the use of and eternal GPS and supports live differential corrections.
They we will fly using Pix4D Capture. It is not open source. It is free and there is a community support forum. Any equivalent software can be used to get the pictures.
We discuss some good practices for the flight, like flight height, image overlapping, camera angle, etc. Some flights will be made available using different configurations.
Then we run OpenDroneMap. OpenDroneMap is a huge software stack. Processing images with full resolution requires lots of RAM and takes time. We will see how easy is to use multiple nodes to run it (so you can take advantage of computers near by). Before using the point cloud in QGIS, we will briefly introduce some nice visualizations tools already available in ODM.
Finally, we will open the point cloud in QGIS.