3D Shipwreck Documentation: Marine Archaeologist Guide

Marine archaeologists use 3D documentation to create precise, measurable digital replicas of shipwrecks, enabling detailed study, monitoring, and public engagement without disturbing the site. This guide covers the essential steps—from planning and data capture to processing and archiving—using photogrammetry and structured light scanning, the most common methods. You’ll need dive gear, a camera or scanner, and specialized software, but the process is accessible to trained professionals and advanced students.

What You’ll Need

• Underwater camera (DSLR or mirrorless with wide-angle lens) or a structured-light scanner (e.g., for close-range work)
• Underwater housing and lighting (strobes or video lights)
• Scale bars and control points (for photogrammetry)
• Dive team (at least two divers for safety and efficiency)
• Photogrammetry software (e.g., Agisoft Metashape, RealityCapture) or scanning software
• 3D modeling/processing software (e.g., Blender, MeshLab)
• High-capacity memory cards and backup storage
• Underwater writing slate and measuring tape
• GPS or acoustic positioning system for site location

Step-by-Step

1. Plan the survey. Define the wreck’s boundaries and the level of detail needed. For large sites, break into sections. Check conditions: visibility, current, and depth. Obtain necessary permits and ensure dive safety protocols.
2. Set up control points. If using photogrammetry, place scale bars and coded targets on the wreck. Measure distances between them with a tape measure for accurate scaling.
3. Conduct a reconnaissance dive. Swim the site to identify key features, hazards, and the best lighting angles. Note any areas that need special attention.
4. Capture overlapping images. For photogrammetry, take a series of photos with at least 60-80% overlap between consecutive images. Move in a systematic pattern: along the length, then across the width, and at different heights. Keep the camera at a consistent distance (e.g., 1-2 meters) and angle. For structured light scanning, slowly sweep the scanner over the surface, keeping a steady speed.
5. Record environmental data. Note visibility, water temperature, and lighting conditions—these affect processing. Also record the GPS coordinates of the site.
6. Back up data on site. After the dive, immediately back up images to a waterproof external drive. Verify that images are sharp and well-exposed.
7. Process photogrammetry. Import images into software. Align photos, build a dense point cloud, then generate a mesh and texture. For structured light, merge scans and clean the mesh.
8. Scale and georeference. Use control points to scale the model to real-world dimensions. If GPS data is available, georeference the model to a coordinate system.
9. Clean and optimize the model. Remove noise, fill small holes, and decimate the mesh if needed for performance. Create a low-poly version for web viewing if required.
10. Archive and publish. Save the raw data, processed model, and metadata in a stable repository. Publish a subset for public access, such as on a museum website or academic platform.

Common Mistakes

• Insufficient overlap between photos, leading to gaps in the model.
• Poor lighting causing blurry or underexposed images.
• Not using scale bars, making the model impossible to measure accurately.
• Ignoring current movement that causes camera motion blur.
• Failing to back up data immediately, risking loss.

Tips

• Use a camera with manual settings to ensure consistent exposure.
• Take more photos than you think you need—extra coverage is better than gaps.
• For large wrecks, break the survey into overlapping sections and merge later.
• Practice on a small object before diving to get comfortable with the workflow.
• Use a dive light at an angle to bring out texture and reduce backscatter.

Troubleshooting

• Alignment fails: Check that images are sharp and have enough overlap. Remove blurry frames and try again. Reduce the number of images if the software struggles.
• Model has holes: Increase photo overlap in those areas or use a scanning pattern that covers more angles. In processing, use hole-filling tools but be aware they are approximations.
• Scale is off: Ensure scale bars are visible in the images and correctly identified in the software. Double-check the measured distances.
• Backscatter (white spots): Adjust lighting to reduce particles in the water, or use a longer strobe arm. In processing, use a dust filter to remove small artifacts.
• Software crashes: Close other programs, reduce image resolution for initial alignment, and save frequently.

Frequently Asked Questions

How long does it take to document a wreck in 3D?

For a small wreck (e.g., a 10-meter boat), a few dives for data capture and a day or two of processing. Larger sites can take weeks to months.

What is the best software for photogrammetry?

Agisoft Metashape and RealityCapture are industry standards. Open-source options like Meshroom are free but may require more tweaking.

Can I use a GoPro for 3D documentation?

Yes, but image quality and distortion can affect accuracy. For professional results, a DSLR with a wide-angle lens is preferred.

How accurate is photogrammetry underwater?

With proper scale bars and good overlap, accuracy can be within a few centimeters. Structural light scanning can achieve millimeter accuracy but requires close range.

Do I need special diving training?

Yes, you should have advanced open water certification and ideally training in underwater survey techniques. Navigation and buoyancy control are critical.

Can I document a wreck without touching it?

Yes, both photogrammetry and structured light are non-contact methods, making them ideal for fragile sites.

Related

• Office Tab 14.00 Release Notes: What’s New
• iPhone 1.6nm Processors? The Truth Explained
• Galaxy S9 Iris Scanner: Does It Have One?