What is photogrammetry?

Photogrammetry is a technique that turns ordinary photos into accurate 3D measurements and models. By taking many overlapping pictures of an object, building, or landscape from different angles, special software can calculate the exact positions of points in space and recreate the shape in three dimensions.

Let's break it down

  • Take pictures: Capture a series of photos that cover the whole subject, making sure each photo overlaps the next by about 60‑80%.
  • Find common points: The software looks for tiny details (like corners or texture) that appear in multiple photos.
  • Triangulate: Using the known positions of the camera for each photo, it draws lines from each camera to the common points. Where the lines intersect is the 3D location of that point.
  • Build a mesh: Connecting all those points creates a wire‑frame surface, which can be turned into a solid 3D model.
  • Add texture: The original photos are wrapped onto the model, giving it realistic colors and details.

Why does it matter?

Photogrammetry lets anyone create precise 3D data without expensive laser scanners or specialized hardware. It makes it possible to document real‑world objects, preserve cultural heritage, plan construction projects, and create assets for games or VR-all from a regular camera or smartphone.

Where is it used?

  • Archaeology & heritage: Scanning statues, ruins, and artifacts for preservation and study.
  • Construction & surveying: Mapping terrain, building sites, and monitoring progress.
  • Film, games, and VR: Generating realistic environments and props quickly.
  • Agriculture: Analyzing crop health and field topography from drone images.
  • Forensics & accident reconstruction: Recreating scenes to understand what happened.

Good things about it

  • Low cost: Only a camera and software are needed.
  • Portable: Works in the field with smartphones or drones.
  • High detail: Captures texture and color directly from photos.
  • Non‑contact: Safe for delicate or hazardous objects.
  • Scalable: From small objects to entire cities.

Not-so-good things

  • Requires good lighting and clear textures; shiny or uniform surfaces can confuse the software.
  • Needs many overlapping photos; missing angles create holes in the model.
  • Processing can be CPU/GPU intensive and time‑consuming for large datasets.
  • Accuracy depends on camera calibration and precise positioning; without proper setup, measurements may be off.
  • Large file sizes for high‑resolution images and resulting models can be hard to store or share.