3D Scanning for Archaeologists: Mapping Hidden Rooms

Updated 5 November 2026 · Standard formula, shown below

3D scanning lets archaeologists map hidden rooms and burial vaults without disturbing them, using lasers or structured light to capture precise geometry through small openings or even through walls. This guide covers the process from choosing the right scanner to processing the point cloud into a usable map, with practical steps for fieldwork and common pitfalls to avoid.

What You’ll Need

A 3D scanner: handheld (e.g., LiDAR or structured light) or tripod-based (terrestrial laser scanner). For tight spaces, a compact handheld unit is best.
Access to the site: a small opening (e.g., a hole in a wall or floor) or a clear line of sight through a doorway.
Targets or markers: reflective stickers or coded targets for aligning multiple scans.
Processing software: e.g., RealityCapture, Agisoft Metashape, or open-source CloudCompare.
Rigid pole or extension arm (for inserting the scanner through small openings).
Laptop or tablet with the scanner’s companion app for real-time preview.
Batteries and memory cards (plenty of spares).
Protective gear: gloves, dust mask, hard hat, and proper lighting.

Step-by-Step

Assess the site and access point. Determine the size of the opening, the depth of the space, and whether it’s safe to enter. If the space is too small for a person, plan to use a pole-mounted scanner.
Choose the right scanner. For rooms you can enter, a handheld LiDAR scanner (like a GeoSLAM or Leica BLK) works well. For tiny openings, use a compact structured-light scanner (e.g., a small USB probe) or a terrestrial scanner placed outside the opening.
Set up targets. If the scanner uses targets for alignment, place them in the visible area or inside the room (if accessible) so they appear in multiple scans. For hidden rooms, use coded targets that the software can auto-detect.
Plan the scan path. For handheld scanning, move slowly and steadily, covering the entire area from multiple angles. For tripod-based scanning, set up the scanner at several positions around the opening to get overlapping views.
Perform the scan. Start the scanner and move it through the space, keeping a consistent distance from walls and floors. For a pole-mounted scanner, insert it through the opening and rotate it slowly to capture the full 360Â° view. Monitor the live preview to ensure you’re not missing areas.
Capture multiple scans if needed. For large or complex spaces, do several overlapping scans. Each scan should share at least 30% overlap with the next for good alignment.
Transfer data to processing software. Import the raw scan data (point clouds) into your software. If using targets, the software will automatically align scans; otherwise, manually align using common features.
Align and merge the scans. Use the software’s alignment tools (e.g., ICP or target-based registration) to combine the scans into a single point cloud. Check for misalignment and adjust manually if needed.
Clean the point cloud. Remove noise, stray points, and any objects that aren’t part of the structure (e.g., debris or the scanner pole). Use filtering tools to smooth the data.
Generate a mesh or model. Convert the point cloud into a triangulated mesh to create a solid surface. This can be exported as an STL or OBJ file.
Create 2D maps and sections. Use the software to slice the model horizontally at different heights to produce floor plans, or vertically for cross-sections. This is what archaeologists use to document the layout.
Export and archive. Save the final point cloud, mesh, and maps in standard formats (e.g., .las, .ply, .dxf). Include metadata like coordinate system and date.

Common Mistakes

Moving too fast. Fast movement creates blurry or sparse point clouds. Slow, deliberate motion is key.
Ignoring overlap. Without enough overlap, scans won’t align properly, leading to gaps or double walls.
Skipping target placement. In featureless rooms (like empty vaults), targets are essential for alignment; relying on natural features may fail.
Not checking the live preview. You may miss entire walls or ceilings, only discovering after you’ve left the site.
Over-cleaning the data. Removing too many points can erase important details like carvings or cracks.

Tips

Always carry spare batteries and memory cards; 3D scanning drains power quickly.
Use a tripod for static scans if you need high accuracy; handheld is faster but less precise.
For dusty environments, use a scanner with IP54 rating or higher to protect against particles.
If the space is very dark, add a light sourceâ€”but avoid shiny surfaces that can cause reflections.
Practice on a known room first to calibrate your scanning speed and settings.
Use the software’s automatic noise reduction, but always double-check the result.

Troubleshooting

Scans won’t align. Try adding more targets, increasing overlap, or using manual alignment with at least three common points.
Point cloud has holes. Re-scan the missing areas, or use the software’s hole-filling tool (but note that this creates estimated geometry).
Data is too noisy. Check for dust, fog, or reflective surfaces. Clean the scanner lens and re-scan if necessary.
Scanner loses tracking. This often happens in featureless spaces. Move slower, or place temporary markers (e.g., tape) on walls to give the scanner reference points.
File size is huge. Downsample the point cloud (e.g., to 1 cm spacing) if you only need a map, not fine detail.

Frequently Asked Questions

Can 3D scanning see through walls?
No, standard 3D scanners use light, so they can’t see through solid walls. However, ground-penetrating radar (GPR) can detect voids, and then 3D scanning is used once an opening is made.

What’s the difference between LiDAR and photogrammetry?
LiDAR uses laser pulses to measure distances directly, working in the dark. Photogrammetry uses multiple photos to reconstruct geometry, requiring good lighting and texture on surfaces.

How accurate is 3D scanning for archaeological mapping?
Accuracy varies by scanner: terrestrial laser scanners can achieve millimeter-level precision, while handheld units are typically within a few centimetersâ€”sufficient for most mapping purposes.

Do I need special software to process the scans?
Yes, you’ll need software that can handle point clouds. Free options include CloudCompare and MeshLab; commercial options like RealityCapture are more automated.

Can I scan a vault that is completely sealed?
No, the scanner needs a line of sight. If there’s no opening, you’ll need to use non-invasive methods like GPR or micro-cameras before scanning.

How long does it take to scan a small room?
A handheld scan of a 3×3 meter room might take 10-15 minutes, plus additional time for processing and cleaning the data.

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.