VFX Atlas
H 292 / 3212010s

Photogrammetry

Photogrammetry / structure from motion. 2018: Meshroom facilitates the open photogrammetry workflow; older methods

Typeface · Space Mono

Meshroom: View of a reconstructed scene in the official user manual.
Meshroom: View of a reconstructed scene in the official user manual.AliceVision / Meshroom Documentation Team.View source ↗

Prompt

Apply a Photogrammetry effect, rendering the scene as a reconstructed three-dimensional scan covered in photographic textures. Introduce subtle triangulated surface irregularities, baked lighting, stretched textures near occlusions, and small rough mesh gaps along complex edges, while retaining realistic color and recognizable volume. Keep the original subject and composition.

Paste it into LetzAI with your own image to apply the effect ↗

Identify

Photographic textures over reconstructed three-dimensional surfaces.

Principle

Many overlapping photos provide camera positions1, depth, and a textured model2.

Example

AliceVision Meshroom, first released in 2018.

Try it out

Photograph a matte object from all sides with strong overlaps.

Background

Photogrammetry extracts geometry from photographs and has a long history predating digital creative tools. This slide anchors the more recent development to the first Meshroom release from August 2018. AliceVision describes its open pipeline of features, camera estimation, dense reconstruction, and texturing. The method requires recognizable image features across multiple shots. Glossy, transparent, or low-texture surfaces complicate this because views do not reliably match. Therefore, a matte, detailed object in even lighting is suitable for this exercise. The surface color is taken from the photographs; the shape is reconstructed computationally. Missing areas and torn edges can later be intentionally left visible as part of the scan aesthetic.

Distinction

Photogrammetry is a measurement and reconstruction method; the scan look is a possible aesthetic use.

Aftermath

A scan is not a neutral impression: the image cover, light and algorithm determine which parts are fully and stably reconstructed.

Further examples · 8 images

Photogrammetry of the face 03
02 Photogrammetry of the face 03A dense point cloud of a human face is visible against a gray background. It demonstrates the initial geometric extraction from photographs before surface texturing is applied.Okhrimenko Roman · CC BY 4.0 · Wikimedia Commons
Wall Painting
03 Wall PaintingA three-dimensional reconstruction shows a section of a wall with graffiti and brick textures. It illustrates how photogrammetry captures both color and shape from real-world matte surfaces.DanielNaseri · CC BY-SA 4.0 · Wikimedia Commons
Sede da Fazenda do Pinhal, N.ELAC
04 Sede da Fazenda do Pinhal, N.ELACA software interface shows a 3D building model alongside the source photographs used for reconstruction. It illustrates the pipeline of matching camera positions to generate a digital architectural twin.NELAC · CC BY-SA 4.0 · Wikimedia Commons
University of Portsmouth CCIXR 3 Photogrammetry & Scanning Studio
05 University of Portsmouth CCIXR 3 Photogrammetry & Scanning StudioA professional studio contains multiple cameras mounted on vertical rails around subjects. This setup allows for the simultaneous capture of overlapping photos necessary for high-fidelity 3D human scans.Tim Sheerman-Chase · CC BY 2.0 · Wikimedia Commons
Esper LightCage photogrammetry camera rig
06 Esper LightCage photogrammetry camera rigA woman sits inside a spherical rig equipped with numerous lights and cameras. This specialized hardware ensures even lighting and total coverage for precise geometric feature extraction.ESPER HQ · CC BY 2.0 · Wikimedia Commons
Photogrammetry model
07 Photogrammetry modelA textured 3D mesh of a landscape feature is shown with XYZ coordinate arrows. It highlights how photogrammetry translates photographic features into a measured three-dimensional spatial model.Razadazs · CC BY 4.0 · Wikimedia Commons
Camera Reconstruction in Meshroom - visualization
08 Camera Reconstruction in Meshroom - visualizationThe interface displays estimated camera positions relative to a reconstructed 3D point cloud in Meshroom. It clearly illustrates the computational process of deriving spatial coordinates from a sequence of photographs.Colin Kranz · CC BY-SA 4.0 · Wikimedia Commons

Question

Why is a shiny glass more difficult to scan than a matte stone?

Show possible answer +

Reflections and transparency change the visible features with the viewing angle; stable correspondences between photos are lacking.

Sources