Drone vs. manual survey vs. satellite
Which method delivers the most accurate roof area with the least time on site? Comparison, accuracy and a glossary of the key terms.
How accurate is a drone roof survey?
With correct image capture (nadir + oblique, 20–30 m above the ridge, 70–80 % overlap), a drone roof survey achieves an area accuracy in the range of a few centimetres: in internal tests the deviations were 0.1–0.6 %. With incomplete coverage, too little overlap or too great a flight height or distance from the roof, the accuracy drops accordingly.
- Combine nadir and oblique shots: nadir for the surface, oblique for edges and eaves.
- Stay 20–30 m above the ridge — higher reduces resolution, lower needlessly increases the image count (guideline).
- Keep overlap at 70–80 % so photogrammetry can reconstruct every point from several images.
- A smaller ground sampling distance (GSD) means finer detail — it depends on flight height and camera.
The figures above are professional guidelines for a well-planned capture; the actual accuracy depends on the roof, weather, camera and image quality.
Drone vs. manual survey vs. satellite
The three common methods compared directly by accuracy, time on site and whether a roof inspection is required.
| Method | Accuracy | Time on site | Inspection needed | Cost per project |
|---|---|---|---|---|
| Drone (photogrammetry) | a few cm with good capture | approx. 10–15 min | No | low |
| Manual survey | centimetres to decimetres, error-prone | 1–2 hours | Yes | medium to high |
| Satellite / aerial imagery | coarse, imprecise for detailed areas | none (instantly available) | No | very low |
Practical guideline values for drone roof surveying. The drone method avoids climbing onto the roof (occupational safety) while delivering the highest level of detail accuracy.
Roof surveying glossary
The key terms around photogrammetry and roof surveying — briefly explained.
- Nadir
- A vertical shot pointing straight down. Nadir images capture the roof surface with little distortion and are the basis for area calculation.
- GSD (Ground Sampling Distance)
- Ground resolution — the distance on the ground covered by one image pixel. A smaller GSD means finer detail; it decreases with lower flight height and higher camera resolution.
- Orthophoto
- A rectified, true-to-scale aerial image. Distances and areas can be measured directly and to scale in an orthophoto.
- Photogrammetry
- A method that computes a 3D model and precise measurements from many overlapping photos. The foundation of drone roof surveying.
- Ridge
- The topmost, horizontal edge of a pitched roof where two roof surfaces meet.
- Hip
- The outward-sloping edge where two roof surfaces meet around a corner (convex edge).
- Eave
- The lower, horizontal edge of the roof where rainwater runs off — usually with a gutter.
- Valley
- The inward-sloping edge where two roof surfaces form a channel (concave edge) in which water collects.
- Overlap
- Consecutive drone images overlap by 70–80 %. This high overlap gives the photogrammetry software enough common feature points to compute a precise model.
- Verge (Ortgang)
- The sloping edge at the gable end of a pitched roof. Unlike the eave (horizontal lower edge), the verge runs diagonally and forms the lateral boundary of the roof surface.
- Parapet (Attika)
- The low masonry wall that terminates the perimeter of a flat roof. It conceals the roof slope and protects against wind and overlooking.
- Orbit flight (oblique and facade orbit)
- A circular flight around the building, supplementing the vertical nadir images. For 2D roof surveying with the camera tilted downward (gimbal approx. −45°) so that edges and eaves are clearly visible. For a 3D model with facades, add an orbit with a horizontal camera (approx. 0°, facing the facade head-on). 70–80 % overlap in each case.
When is drone surveying worth it instead of a manual survey?
Drone surveying is worth it as soon as accuracy, occupational safety and time on site matter — that is, for almost any roof in solar, roofing and planning projects. It avoids climbing onto the roof, fully documents the current state and delivers a reusable 3D model instead of individual hand measurements.
Calculating roof area from drone images
Roofy calculates the roof area fully automatically from your drone images: no manual drawing or roof access required. Photograph the roof from 20–30 m with nadir and oblique images (70–80 % overlap), upload the images, and the system delivers areas and edges as a PDF report.
- Nadir images (camera pointing straight down) capture roof surfaces without distortion.
- Oblique images (gimbal −45°, orbit flight) add façades and edges for a complete 3D model.
- With 70–80 % overlap, Roofy delivers reliable measurements for solar installations, waterproofing, and planning.
Reference project: detached house Galgenen (floor plan 13.92 × 10.12 m per architectural drawing) from 82 drone images.
Next step
How to measure a roof in Roofy
Draw a polygon, check the elevation map and set measurement lines — step by step.