# Choose settings for your objective

> Orthophoto, earthworks, volumes, assets, vegetation or 3D: what to capture and which options matter.

Page: https://cloudodm.stellaris.cl/learn/objectives

## A map for GIS or planning

Capture: nadir photos, 70–80% overlap, height set by the ground sample you need (about 2–3 cm at 80–100 m on most drones). Options: Orthophoto resolution to match your ground sample; “Orthophoto only, fastest” if nothing else is needed. Check: open it on a map in Results and fade it against the street map.

## Terrain, contours and earthworks

Capture: overlap as above, plus ground control points if absolute heights matter. Options: Classify ground, Terrain model, Contour lines, Elevation model resolution about twice the ground sample. For LiDAR the Automatic ground filter is right for almost every site.

What we measured: SMRF alone kept a stadium's stands and large buildings in the terrain model; CSF alone removed them but shaved steep slopes (27% of a mountain site). Automatic combines them. A terrain model removes what stands on the ground, not what the ground was shaped into — embankments and earth-built stands stay, and should.

## Stockpiles and volumes

Capture: 70%+ overlap and some oblique photos around tall piles, so their sides are reconstructed. Options: Surface model, Point cloud detail at least Medium. Measure: in the desktop viewer draw the toe of the pile with the Volume tool and choose the base — lowest point, mean, toe plane or a surface — to get cut and fill.

## Buildings, assets and inventories

For LiDAR, the deep-learning tier is the one that gets buildings right (83% vs 19%) and adds vehicles, poles and fences. Correct what remains in the desktop viewer with the class brush, then export the classes you need.

## Vegetation and canopy

The height of vegetation is the surface model minus the terrain model. Produce both, then hover in the desktop viewer's map view: it reads the surface, the terrain and the height above ground under the pointer. LiDAR sees through gaps in the canopy; photogrammetry only sees the top.

## A 3D model to show

Capture: add oblique passes around the object. Options: “Full 3D mesh” and “3D model for the web”. The model opens in the Results viewer and as GLB in any 3D tool.

Every option in the composer says what it does and what it costs. When in doubt, keep the default and look at the result before paying for more detail.

## More about CloudODM

- [CloudODM: photogrammetry and LiDAR in the cloud](https://cloudodm.stellaris.cl/)
- [Cloud photogrammetry](https://cloudodm.stellaris.cl/fotogrametria-en-la-nube)
- [LiDAR classification](https://cloudodm.stellaris.cl/clasificacion-lidar)
- [Point cloud viewer](https://cloudodm.stellaris.cl/visor-nube-de-puntos)
- [Pricing](https://cloudodm.stellaris.cl/pricing)
- [Features](https://cloudodm.stellaris.cl/features)
- [Measured accuracy](https://cloudodm.stellaris.cl/accuracy)
- [Demonstration: a finished survey](https://cloudodm.stellaris.cl/demo)
- [How it works](https://cloudodm.stellaris.cl/how)
- [Tutorials](https://cloudodm.stellaris.cl/learn)
- [ODMDisplay, the desktop application](https://cloudodm.stellaris.cl/viewer)
- [Contact](https://cloudodm.stellaris.cl/contact)
- [Terms and privacy](https://cloudodm.stellaris.cl/terms)