We have plans for measuring a number crop ecophysiological properties using three different cameras mounted on a multi-rotor helicopter drone (UAV) - digital RGB, thermal, and multispectral cameras. In addition to a variety of vegetation indices, we would also like to add the capacity of estimating plant height using overlapping images and the on board GPS capabilities. I have done this sort of analysis some time ago with satellite data, but am not currently aware of the open-source options for this analysis.
It's a very interesting project!
Maybe to estimate the real measure you can use two or more laser pointers at fixed distance mounted on one camera (or more): something like in exploration under the sea (https://forum.openrov.com/t/laser-pointers/631/30) or instead you can try to use a calibrated camera (such as it is used in photogrammetry).
Hope this helps,
Best regards
Francesco
If you take multiple images of your area of interest you could use the Structure-from -Motion approach to model the veg in 3D (similar to the link Anupam posted). Just be aware that if you rely solely on the onboard GPS for georeferencing your error bars could be rather large. Two open source SfM softwares are:
http://ecosynth.org/
http://ccwu.me/vsfm/
Thanks for the answers and suggestions! Thanks, James, I will have to check out those two open source SfM softwares. We have been flying Mikrokopter drones over wheat in Spain with great results so far (still processing the data sets). I have been using VisualSfM open source and Agisoft (commercial trial license). They are fairly comparable in results, with the commercial option offering some nice extras. For the most part I am using ground-based GPS to register the models after combining the cameras. We just made Spanish national TV in case y'all are interested the link is on my professinoal blog at http://ecostem.org/news-blog/
- Badges
- Science topic
- Similar topics
- Geoscience
- Geoinformatics
- Geographic Information System