A 3D scanner captures surface geometry — but what exactly does that mean in practice? What information makes it into the scan data, what gets left out, and what happens to that data once the scanner is put away? These are practical questions that matter before you commit to a scanning project, and the answers help set realistic expectations for what the data will and won’t let you do.
The short answer: a 3D scanner captures the shape and position of an object’s exterior surfaces in three-dimensional space. It does not capture interior geometry, material properties, weight, color (unless specifically configured to), or anything that can’t be directly measured by light reflecting off a surface.
Points, Meshes, and What the Raw Data Looks Like
The fundamental output of a 3D scan is a point cloud — a dense collection of individual coordinate measurements, each representing a position on the object’s surface in 3D space. Depending on the scanner and the object size, a point cloud might contain anywhere from tens of thousands to hundreds of millions of points.
Point clouds are then processed into mesh files — surface models made up of triangles that connect the measured points into a continuous surface. The most common formats are STL and OBJ. These files can be opened, visualized, and used as reference geometry in most CAD and engineering software.
The mesh represents the object as it physically existed at the time of scanning: every contour, every surface irregularity, every dimension as measured. It’s a snapshot of physical reality — which is both its greatest value and its primary limitation.
What 3D Scan Data Captures Well
3D scanners excel at capturing:
- Complex organic geometry — Freeform curves, sculptural surfaces, and irregular contours that would be extremely difficult to measure manually or define parametrically from scratch.
- As-built conditions — The actual current state of a part, assembly, or structure — including wear, deformation, or modifications made since original manufacture.
- Fine surface detail — Engraving, texture, stamped markings, and subtle relief features, depending on the scanner’s resolution.
- Large and geometrically complex assemblies — When individual components need to be captured in their installed context, scanning provides spatial relationships that would be impractical to measure any other way.
- Spatial relationships — How surfaces relate to each other in three-dimensional space, including features that are awkward to reach with physical measurement tools.
What 3D Scan Data Does Not Capture
Understanding the limits of scan data is just as important as knowing its strengths:
- Interior geometry — Standard optical scanners only see what light can reach. Internal channels, blind holes, undercuts, and enclosed cavities may be partially or completely absent from scan data unless specialized approaches (CT scanning, for example) are used.
- Material properties — Scan data carries no information about what the object is made of, its hardness, its thermal properties, or anything beyond surface geometry.
- Engineering intent — The scan captures what is, not what was designed to be. If a part has worn, deformed, or been modified, those conditions are captured too. The scan has no way of knowing what the part was originally supposed to look like.
- Color and texture (by default) — Geometry scanners capture shape, not appearance. Color scanning is possible with photogrammetry or specialized hardware, but it’s a separate data stream and not standard on most engineering-grade systems.
- Tolerances and parametric features — Scan data is a mesh, not a CAD model. There are no defined tolerances, no editable features, no design parameters embedded in the file.
From Scan Data to Usable Engineering Output
Raw scan data is a starting point, not a finished deliverable for most engineering applications. What happens next depends on how the data will be used.
For reverse engineering, the mesh becomes reference geometry for rebuilding a clean, parametric CAD model — one that has defined features, tolerances, and can be modified and sent to manufacturing. This is the conversion from “a record of what exists” to “a model that defines what to make.”
For inspection and quality control, scan data is compared against an existing CAD model or drawing to generate a deviation analysis — a color-mapped report showing where manufactured parts fall within or outside tolerance.
For archival or visualization purposes, the mesh itself may be the final output — a high-fidelity digital record of an object, useful for documentation, rendering, or future reference without necessarily needing further processing.
At Kemperle Industries, we work with clients at every stage of this pipeline — from raw capture through to engineered CAD models ready for fabrication. If you’re figuring out what to do with scan data, or whether scanning is the right first step for your project, we’re happy to talk it through.
