An orthomosaic looks like a satellite image but better: sharper, more current, and at centimetre resolution. It is built by stitching hundreds of overlapping drone photos and correcting each one for lens distortion, terrain and camera tilt, producing a single map where distances can be measured accurately. For surveyors, farmers and site managers, it is often the most useful product a drone can produce.

The difference from a simple stitched panorama is geometric correction. A raw aerial photo distorts scale: objects near the edges appear stretched, tall objects lean outward, and terrain height shifts positions. Orthorectification uses the terrain model built during photogrammetric processing to correct every pixel to its true ground position, so the final mosaic behaves like a map, not a photograph.

Capture quality decides processing success. Fly a grid pattern with at least 75 percent front overlap and 70 percent side overlap, keep the camera pointing straight down for flat sites, and shoot at full resolution with a fast shutter speed to freeze motion blur. For sites with buildings or stockpiles, add oblique shots tilted to about 45 degrees from several directions so the software can reconstruct vertical surfaces.

Ground control points turn a good-looking mosaic into an accurate one. Place at least five marked targets across the site, measure their coordinates with a survey-grade GPS receiver, and include them in processing. Without ground control, the mosaic may be offset by metres from its true position, which matters for surveying, construction and any work that overlays the mosaic with cadastral or design data.

Processing options range from free to enterprise. OpenDroneMap is a capable open-source pipeline that runs locally or in the cloud and produces orthomosaics, elevation models and point clouds. Commercial tools like Pix4D, Agisoft Metashape and DJI Terra offer polished interfaces and faster processing at a price. Cloud services process on remote servers, which suits users without powerful computers. All follow the same stages: align photos, build a dense point cloud, build a surface model, then orthorectify and mosaic.

Loading the result in QGIS is straightforward: add the exported GeoTIFF as a raster layer and check that its coordinate reference system matches your project. Verify accuracy by comparing the mosaic against known features, such as surveyed points or a trusted basemap, and note any offset before using it for measurements. Set the layer's rendering to show true colour and adjust contrast if the mosaic looks washed out.

Common failures have familiar causes. Insufficient overlap produces holes or warped areas. Moving subjects, such as vehicles or water, confuse matching. Strong shadows and reflective roofs create artefacts. Flying at midday with the sun high reduces shadows, and avoiding windy days keeps vegetation still, both of which improve stitching noticeably.

An orthomosaic workflow is a skill that compounds: each flight teaches you something about overlap, light and ground control that makes the next map better. Start with a small flat site, verify the result carefully, and build toward larger, more complex projects with confidence grounded in checked outputs.

Before delivering an orthomosaic to a client, run a final quality pass that many beginners skip: check the edges of the mosaic for stretching, verify a few known distances against ground measurements, and confirm the coordinate system in the file metadata matches what the client expects. That ten-minute check is the difference between a deliverable and a callback. Over time, a library of verified orthomosaics becomes one of the most valuable assets a surveyor or site manager owns, because it records conditions exactly as they were on a known date.