Structure from Motion
Collect Drone data
Fly the drone
Tips for mapping: * https://help.dronedeploy.com/hc/en-us/articles/1500004964282-Making-Successful-Maps
SfM
typical data flow
Once finished, download all assets
transform tiff to COG and push to S3 bucket
Transform entwine point cloud to potree
Scaling ODM
ODM is a docker image, so in theory it could be put on AWS, push the images to a bucket, spin up a lot of instance for the processing and scale back when done. I would like to know how to execute this but at the moment I can get all my processing done here so there’s no reason to invest more time in it now.
SFM in adverse conditions
- Meshlab
Collection
Processessing
These come out as abc or ply, neither of which are currently supported formats for Potree. Therefore, we’ll need to convert them which leads us to:
Potree converter
https://github.com/potree/PotreeConverter https://github.com/potree/PotreeConverter/releases
Install
Run
```{Powershell}
.\PotreeConverter.exe G:\dronedata\KS_LearnedFlume\sedflume\20210310\meshroom\MeshroomCache\ConvertSfMFormat\c2ffbee2f8065999d1a8e24503cb345094f673f4\sfm.ply -o G:\dronedata\KS_LearnedFlume\sedflume\20210310\meshroom\PotreeConvert
```
Others in the space
https://www.unr.edu/uas-federation