A drone survey captures 200 overlapping photographs of a construction site. Structure-from-Motion (SfM) processing produces a 3D point cloud, orthomosaic, and DEM. What is the minimum geometric requirement for SfM to work?
AAll photographs must be taken from exactly the same altitude
BEach point in the scene must be visible in at least two photographs taken from different positions, providing parallax for triangulation
CThe drone must fly in a perfectly straight line
DAt least one photograph must include a GPS antenna for georeferencing
SfM requires overlapping photographs from different viewpoints so that feature matching algorithms can identify common points across images and compute their 3D positions through triangulation. The parallax (apparent shift of points between views) encodes depth information. In practice, 60-80% forward overlap and 30-60% side overlap between flight lines ensures every ground point appears in multiple images.
Question 2 True / False
Photogrammetric DEMs derived from optical imagery can see through forest canopy to map the bare-earth surface, just like LiDAR.
TTrue
FFalse
Answer: False
Photogrammetry matches visible surface features between images, so it maps what is visible from above -- the tops of trees, not the ground beneath them. In forested areas, photogrammetric DEMs are Digital Surface Models (DSMs), not bare-earth DTMs. LiDAR pulses physically penetrate canopy gaps to reach the ground. This is a fundamental limitation of passive optical 3D reconstruction compared to active laser-based measurement.
Question 3 Short Answer
Explain what ground control points (GCPs) contribute to a photogrammetric survey and when they might be unnecessary.
Think about your answer, then reveal below.
Model answer: GCPs are points with precisely known coordinates (typically surveyed with GPS) that are visible in the photographs. They anchor the photogrammetric model to real-world coordinates, correcting for systematic errors in camera position and orientation. They also enable accuracy assessment by comparing photogrammetric coordinates against known values. GCPs may be unnecessary when high-accuracy RTK/PPK GPS is integrated with the drone, providing centimeter-level camera positions directly, or when absolute georeferencing is not required (e.g., relative measurements within a scene).
GCPs transform an internally consistent but arbitrarily oriented 3D model into a georeferenced product with known accuracy in a real-world coordinate system.