Knowledge centerLaser Scanning vs. Photogrammetry
Laser scanning and photogrammetry both end in a point cloud, which is why they are often treated as interchangeable. They are not. One measures distance directly; the other calculates geometry from photographs.
The difference in one line
A laser scanner measures where a surface is. Photogrammetry infers where a surface is, from how it appears across overlapping photographs.
What that means in practice
- Light. Scanning works in an unlit basement. Photogrammetry needs usable light.
- Surface texture. Photogrammetry struggles on blank white walls and glass, because it needs visual features to match between frames. A scanner does not care.
- Scale. A scan is dimensionally correct by definition. A photogrammetric model needs control points or a known reference to be scaled correctly.
- Reach. Photogrammetry from a drone covers roofs and facades far faster than a tripod-mounted scanner can.
Why most projects use both
On a typical building we scan the interior, fly the roof and facades, and register both into the same coordinate system. The point is not to pick a winner but to use each where its weakness does not matter. A Matterport sits at the photogrammetric end of that scale, with the accuracy that implies. More terms are defined in the knowledge center.
Which one for which job
| Situation | Method | Why |
|---|---|---|
| Interior, occupied office | Laser | Works in any lighting, no texture required, and the dimensions will be built against. |
| Plant room, ceiling void | Laser | Unlit, dense, and photographs cannot be taken at a useful angle in the space available. |
| Roof and upper facade | Photogrammetry from the air | Reaches what a tripod cannot, and covers a whole roof in one flight. |
| Large open site, stockpiles | Photogrammetry from the air | Area per hour. Terrestrial scanning would take days for the same coverage. |
| Ornament, moldings, carving | Either, often both | Laser for geometry, imagery for surface detail and for anything that has to be replicated visually. |
| White membrane roof, glazing | Laser from the ground | Photogrammetry needs visual features to match between frames and fails quietly where there are none. |
Combining them without losing the tolerance
On most buildings both methods appear in the same project, and the joint between them is where accuracy is won or lost. Aerial capture that is not tied to ground control produces a model that looks right and measures wrong, and the error is invisible until somebody builds against it.
The fix is unglamorous: control points measured on the ground, the flight registered against them, and the whole thing sitting in the same coordinate system as the interior survey. Done properly, an elevation cut through the facade is dimensionally consistent with the floor plan behind it. Done carelessly, the roof floats somewhere near the building.
We also record which method produced which part of the deliverable. A single accuracy figure covering both is convenient and misleading, which is why a USIBD Level of Accuracy is stated per area rather than per project.
Three things people get wrong
- “A point cloud is a point cloud.” Both methods produce one, and the file formats are interchangeable. What differs is how each point got there: measured, or calculated from images. The format tells you nothing about that, which is why the method record travels with the survey.
- “Photogrammetry is the cheap option.” Per square foot of roof, yes. Per plant room, no, because it needs light, texture and angles that a mechanical space does not offer. Cost follows suitability, not sensor type.
- “More photos means more accuracy.” Beyond adequate overlap, more images mainly means more processing. Scale comes from control, not from image count, and no amount of coverage fixes a model that was never scaled.
The instrument side of this is covered in terrestrial scanning vs. mobile mapping, and the underlying measurement in LiDAR.
Frequently asked questions
Is drone data always photogrammetry?
In building documentation, usually yes: the drone carries a camera and geometry is calculated from overlapping images. Airborne LiDAR exists and is used on infrastructure and terrain work, where it is the better tool for vegetation and large corridors.
Can photogrammetry be millimeter-accurate?
Under controlled conditions, at close range, on a well textured object, it can be very accurate. Across a building, in the field, it is dimensionally accurate rather than millimeter-accurate, and we use that wording deliberately.
Which one is cheaper?
It depends entirely on what is being captured. Aerial photogrammetry is far cheaper per square foot of roof. Terrestrial scanning is far cheaper per usable plant room, because photogrammetry frequently cannot produce one at all.
Do both deliver the same file formats?
Yes. E57 and RCP for the registered cloud either way, plus orthomosaics from aerial work. The formats are identical, the provenance is not, and the deliverable states which is which.
Documenting a building?
Tell us what you need measured and we will come back with the method and the deliverables that fit.