What is photogrammetry?

Photogrammetry is the science of measuring and reconstructing the world from photographs. In practice, software finds the same features in overlapping frames, infers camera positions, and builds a mesh or a dense point cloud of the scene. Surveyors, VFX teams, and phone scanning apps all use some version of it.

What it needs

It needs many views of the same space, with enough overlap and texture to lock. A single listing photo does not give you photogrammetry, no matter how good the restyle is. LiDAR on some phones can substitute or assist, but that is still a capture walk, not an upload of one JPEG.

What Roomersive is not

Roomersive does not photogrammetrically reconstruct your house from one picture. It restyles the still (layout stays, materials and furniture change) and generates a 360° / walkable scene of that room. The 3D you look around in is a generated interior consistent with the restyle, not a survey of the drywall.

That is why a Polycam vs Roomersive comparison is a category split: Polycam (and Matterport) are capture; Roomersive is proposal.

Example

You walk a kitchen slowly with a scanning app. Photogrammetry (or LiDAR + photos) can give you a mesh you might measure a counter on. You instead upload the MLS kitchen photo, restyle it to Japandi, and open a walkthrough. Useful for a buyer. Not a measurement workflow. Do not promise the second is the first.

Related terms

• Gaussian splatting — a rendering representation; can be built from capture or from generated views
• Digital twin — industrial meaning vs. marketing meaning
• Panorama — wide image; not a reconstruction
• LiDAR — active depth on some devices; a different capture path than pure photo overlap