Drone mapping turns hundreds of aerial photos into accurate maps and 3D models. It is used in construction, mining, agriculture, land surveying and disaster response. The technique behind it is called photogrammetry. In this guide we explain how drone mapping works step by step, from planning the flight to the final map, in simple terms.
What is photogrammetry?
Photogrammetry is the science of measuring things from photographs. When the same point on the ground appears in several overlapping photos taken from different positions, software can calculate exactly where that point is in 3D space. Repeat this for millions of points, and you get a detailed 3D model of the area.
Your eyes do something similar: each eye sees the world from a slightly different position, and your brain uses the difference to judge depth.
Step 1: Plan the flight
Mapping flights are planned in software, which then flies the drone automatically. The key settings are:
- Area: the boundary of the site to map.
- Flight height: sets the image detail (ground sampling distance, GSD). Lower flights give more detail but take longer. See the GSD formula in our mapping drone buyer’s guide.
- Overlap: how much neighboring photos overlap. Typical values are about 70–80% front overlap (along the flight line) and 60–70% side overlap (between lines). More overlap helps with trees, crops and uniform surfaces.
- Pattern: a simple grid for flat maps, a double grid or angled (oblique) photos for 3D models of buildings.
Step 2: Set up ground control (if needed)
Ground control points (GCPs) are marked targets on the ground whose positions are measured precisely with survey equipment. They tie the map to real-world coordinates. With an RTK or PPK drone, fewer GCPs are needed, but placing a few checkpoints to verify accuracy is still good practice.
Step 3: Fly and capture images
- Fly in good, even light. Overcast days often give excellent results with soft shadows.
- Use a fast enough shutter speed to avoid motion blur.
- Avoid strong wind, which tilts the drone and changes the overlap.
- Keep the camera settings consistent across the whole flight.
- Check a few photos on site before leaving, so you do not discover problems back at the office.
Step 4: Process the images
Photogrammetry software processes the photos in several stages:
- Feature matching: finds the same points in overlapping images.
- Camera alignment: works out where each photo was taken and how the camera was pointing.
- Point cloud: creates millions of 3D points describing the surface.
- Mesh and textures: builds a 3D surface model and covers it with the photo colors.
- Outputs: generates orthomosaics, elevation models and reports.
Processing can take from minutes to many hours depending on the number of images and the computer. Cloud processing services are an alternative to a powerful local computer.
The main outputs
| Output | What it is | Used for |
|---|---|---|
| Orthomosaic | One large, distortion-corrected aerial image | Measuring distances and areas, site plans |
| Digital Surface Model (DSM) | Height of everything, including buildings and trees | Volumes, obstacles, line-of-sight |
| Digital Terrain Model (DTM) | Height of the bare ground | Drainage, earthworks, contours |
| Point cloud | Millions of 3D points | Detailed measurements, CAD and BIM |
| 3D textured model | Realistic 3D model | Visualization, inspection, presentations |
| Contour lines | Lines of equal height | Engineering and planning |
How accurate is drone mapping?
Accuracy depends on the camera, flight height, overlap, positioning method and ground control. Relative accuracy (measurements within the map) is usually very good. Absolute accuracy (position on the Earth) is typically around a meter or more with standard GPS and no GCPs, and a few centimeters with RTK/PPK or good ground control.
Common problems and fixes
- Holes or distortions over trees, water or crops: increase overlap and fly higher.
- Blurry images: increase shutter speed or slow the drone down.
- Map shifted from reality: add ground control or use RTK/PPK.
- Warped edges: extend the flight area beyond the site boundary.
Frequently asked questions
How many photos does a drone map need?
It depends on area, height and overlap: from a few dozen for a small plot to thousands for a large site.
What is an orthomosaic?
A single aerial image created from many photos and corrected so that scale is uniform, allowing you to measure distances and areas directly.
Do I need ground control points?
For high absolute accuracy without RTK/PPK, yes. With RTK/PPK you need fewer, but checkpoints are still recommended.
Which drone is best for mapping?
It depends on site size and accuracy needs. See our fixed-wing vs quadcopter comparison.