orthomosaic-processinglisted
Install: claude install-skill neblihq/drone-skills
# Orthomosaic Processing with OpenDroneMap
Turn a folder of overlapping drone photos into a georeferenced orthomosaic —
plus an elevation model and a 3D point cloud from the same run — using
OpenDroneMap (ODM), which is free, open source, and runs locally through Docker.
The whole workflow is: check the photos are processable, run one Docker
command, wait, then open the results. Most failed runs die for reasons that
were visible in the photos before processing started, so do the checks first —
they take a minute and can save an hour.
## What you need
- **The photos from one flight, in one folder.** JPGs straight off the drone.
Don't mix flights from different days or altitudes in one run unless they
were flown as a single mission.
- **Proper overlap.** Photogrammetry needs each ground point visible in
several photos: roughly 75% front overlap and 65% side overlap is the
standard target for mapping missions. Photos from a casual flight (a few
snapshots, video stills, orbiting one object) usually cannot make an
orthomosaic — that's the single most common cause of failed runs.
- **Geotags help a lot.** Nearly all drones (any DJI, Autel, Skydio, Parrot)
write GPS positions into the photo EXIF automatically. ODM can run without
them, but the result won't be placed on the map.
- **Docker installed** ([docker.com](https://www.docker.com/products/docker-desktop/)),
with at least 8 GB of RAM available to it — 16 GB is comfortable for a few
hundred photos. On Do