Related Experiment Video
Updated: Jun 19, 2026
07:42
In-vivo Detection of Protein-protein Interactions on Micro-patterned Surfaces
Published on: March 19, 2010
Ink detection from surface topography of the Herculaneum papyri.
Giorgio Angelotti1, Federica Nicolardi2,3, Paul Henderson2,4
1Vesuvius Challenge, San Francisco, USA. giorgio@scrollprize.org.
Scientific Reports
|June 17, 2026
Summary
Researchers developed a deep-learning method to read carbonized Herculaneum papyri by analyzing surface topography. This approach detects ink on papyrus using high-resolution 3D imaging, overcoming limitations of traditional X-ray methods.
Area of Science:
- Digital Humanities
- Materials Science
- Computational Imaging
Background:
- Herculaneum papyri present reading challenges due to carbonization of both scrolls and carbon-based ink.
- Traditional X-ray methods struggle with ink detection on papyrus due to minimal attenuation contrast.
Purpose of the Study:
- To investigate the potential of surface morphology for distinguishing inked regions from papyrus.
- To develop and evaluate a deep-learning model for ink detection based on 3D topography.
Main Methods:
- Training a deep-learning model using 3D optical profilometry data from mechanically opened Herculaneum papyri.
- Analyzing the signal within surface topography for ink detection.
- Quantifying the impact of lateral sampling resolution on model performance.
Main Results:
- High-resolution topography alone provides a usable signal for ink detection on the analyzed papyrus dataset.
- Ink detection is achievable without relying on general bulk relief or universal roughness cues.
- Segmentation performance decreases with reduced lateral resolution, indicating the importance of characteristic spatial scales.
Conclusions:
- Surface morphology analysis, particularly high-resolution topography, offers a viable method for ink detection on carbonized papyri.
- Deep learning models can effectively leverage morphological data for reading ancient texts.
- Findings inform resolution requirements for imaging closed scrolls using morphology-based techniques.
