Segments

Segmentation involves mapping sheets of papyrus inside a 3D X-ray scroll volume or fragment volume.

Illustration of Volume Cartographer with an algorithm extrapolating in 3D.

The resulting 3D structure is called a “mesh”, capturing the complex shape of the papyrus sheet.

Some segment meshes from Scroll 1.

The mesh is used to isolate and extract a thin "surface volume" of the mapped papyrus surface in micro-CT. The surface volume is flattened into a stack of 65 image files (layers), where 32.tif is nominally the papyrus surface. 64.tif is “atop” of the surface and 00.tif is “below”.

A flattened surface volume is produced starting from a mesh.

Scrubbing through layers of the surface volume of segment 20230827161846.

The flattened surface volume can be used directly to look for ink, either by visual inspection or with machine learning tools.

Machine learning ink detection models finding a Pi.

Our Segmentation Team has been mapping select regions of the scrolls. The community has made this a significantly more automated process with improved tools, but it still involves considerable human input.

Total segment area created over time by our team and community.

You can view the meshes of our Scroll 1 segments in Volume Viewer (hosted here) and Segment Viewer (hosted here).

For more technical details about how the segmentation team operates, check out this doc: The Segmenter’s Guide to Volume Cartographer.

Data format

You can find all segment data on the data server in the /full-scrolls/ folder.

• Scroll 1 (PHerc. Paris. 4): /full-scrolls/Scroll1/PHercParis4.volpkg/paths/
• Scroll 2 (PHerc. Paris. 3): /full-scrolls/Scroll2/PHercParis3.volpkg/paths/
• Scroll 3 (PHerc. 332): /full-scrolls/Scroll3/PHerc332.volpkg/paths/
• Scroll 4 (PHerc. 1667): /full-scrolls/Scroll4/PHerc1667.volpkg/paths/

There are many different types of files provided in each paths/<id>/ folder:

• meta.json: Metadata of the segment.
• layers/{00-64}.tif: Surface volume of 65 layers.
  • This is most useful for detecting ink.
• <id>.tif: Texture of the surrounding voxels (the maximum of a few of the surrounding layers).
• <id>.obj: Mesh of the segment.
• <id>_points.obj: Pointcloud.
• <id>.ppm: Per-pixel map, a custom data format mapping points between the surface volume and the original 3D volume of the scroll.
• pointset.vcps: Pointset of a segment. More information here
• author.txt: Name of the author of the segment.
• area_cm2.txt: Total surface area, in cm².