Drone photogrammetry quality checks

Blurred images, inconsistent exposure, missing overlap and weak capture geometry can limit the output before processing begins. Check that the image set covers the agreed area and that camera, capture date and position information are preserved. Identify water, reflective surfaces and moving objects that may reconstruct poorly.

An attractive model can still contain positional errors. Inspect the alignment and control report, then assess independent checkpoints that were not used to fit the solution. Compare the achieved result against the project acceptance criteria rather than a generic accuracy claim.

A surface model includes visible objects. A terrain model requires appropriate ground classification and sufficient evidence of the ground. Ask how vegetation, stockpile bases, buildings and occluded areas were treated before deriving contours or quantities.

Validate coordinate reference, vertical datum, units, scale and file compatibility in the receiving application. Review a small handover package with the end user before processing the full programme. Record limitations beside the deliverable so the next person does not mistake a visualization for a validated survey.