3D Scanning & Photogrammetry

Chapter 1: The Photogrammetry Pipeline: What You Need to Know Before You Start

You have, sitting on your bench, an object that matters. Maybe it's a carburetor float bowl from a 1967 Triumph, cracked and repaired once with epoxy that's now failing. Maybe it's a Nixie tube display you want to document before you touch a single wire. Maybe it's a casting from a pattern that no longer exists. Whatever it is, you need to know its shape to within a fraction of a millimeter, and you need that information in a form you can measure from, model against, or replicate.

Photogrammetry is how you get that shape. It is not magic. It is not a black box. It is a deterministic process that takes a set of overlapping photographs and, through a series of well-understood computational steps, produces a three-dimensional surface. The difference between a photogrammetry project that works and one that produces a lumpy, hole-ridden disaster is not luck. It is the disciplined application of capture technique, and it is the focus of this book.

Here is the core truth you must internalize before you take your first photograph: Photogrammetry is a measurement technique, not an imaging technique. Your camera is not being used to take pictures that look good. It is being used to record the angular relationship between thousands of distinct surface features. Every blurry frame, every overexposed highlight, every shadow that obscures a feature is a lost measurement. The algorithm cannot invent data that was