OpenDroneKitWhat can it measure?
Area, perimeter, volume and change, computed against the real DSM and DTM. Volume is verified to 0.00 m³ error against an analytic test surface source. Cut and fill are reported against the DTM, a plane and a lowest-point reference source.
2-D defects found in photos are back-projected onto the reconstructed surface to become georeferenced polygons with square-metre areas source.
What does a measurement look like?
The homepage shows a polygon measured on the Aukerman Park orthomosaic: 23,591.8 m² planimetric area, 609.72 m perimeter, 10,182 m³ above terrain over the 17,137 m² where the 1.23 m/px DSM has data. Those figures come from the development-branch run of the public dataset.
What if the rasters are missing?
Measurements need georeferenced rasters, so they are only produced by the COLMAP engine. Without them the report states that measurements are absent rather than printing zeros: an area in square metres means nothing unless something measured it source.
Ponding and deformation
Two measurements are deliberately not AI. Ponding finds closed depressions in the DSM, reported as area, depth and volume with the survey’s vertical accuracy carried through; it measures where water can collect, not whether water is there now. Deformation compares two surveys and reports only movement above a detection floor built from both surveys’ accuracies FEATURES.md.
Related
Other parts of the pipeline: Mission planning · Flight over MAVLink · Photogrammetry · AI defect detection · Reports · Offline and private