Vedi qualcosa da migliorare? Proponi una modifica.
Conservators need to know what an old document is made of — its inks, pigments and paper — and how past or future treatments affect it. Taking a physical sample damages the object. Non-invasive imaging avoids that, and it matters for preservation at every level, national, state and local.
With grant funding, the Library of Congress's Preservation Research and Testing Division (PRTD) set out to extend its existing hyperspectral imaging system with two more techniques, making one multi-purpose instrument for heritage science. Principal investigator Dr. Fenella G. France led the work with Meghan Hill and Dr. Bill Christens-Barry of Equipoise Imaging LLC.
Why combine techniques
| Technique | What it reveals |
|---|---|
| Spectral imaging (reflectance and transmission) | how a material reflects or passes narrow bands of ultraviolet, visible and infrared light — clues to its makeup, condition and how it was made |
| Fluorescence | many substances absorb light at one wavelength and give it back at a longer one; the pattern can help identify them |
| Polarized light | about 90% of solid substances respond differently to light depending on its orientation, changing brightness or color as they're rotated |
Putting all three in one system means less handling and less light exposure for fragile objects, and no need to take samples.
Building it
- Polarizing and fluorescence equipment from MegaVision Inc. was installed on August 11, 2010; delays in getting it pushed the project's completion to September 2011.
- The system in PRTD's lab in the Madison Building gained a filter wheel with red, green and blue emission filters, so that images could record only the light the filters let through.
- The team built a reference collection of samples to help characterize materials as they worked out protocols.
How the images were made
- A standard sequence first: 13 images, each lit from both sides by a single-wavelength LED at about 45°, plus 4 "raking" images lit from one side at about 75°.
- Then fluorescence images: the object lit by one LED color while a red, green or blue filter sat in front of the lens — using only filters that pass longer wavelengths than the light shining on the object, since fluorescence always comes out at a longer wavelength than goes in.
- Exposures were tuned to fill the camera's range but mostly capped at 10 seconds, to keep light on real collection items within safe limits.
- Images of a white reference sheet at each wavelength corrected for uneven lighting; software removed camera noise and converted files to standard TIFF, and new analysis modules were written for the open-source ImageJ package.
The settings were saved as a standard protocol for future fluorescence imaging.

Transmission spectra of the Wratten emission filters, with LED peak wavelengths shown as dashed lines. Library of Congress
Separating glow from leakage
A filter doesn't block the illuminating light perfectly; some leaks through and could be mistaken for fluorescence. Knowing how much each filter passes at each LED wavelength lets the team calculate how much of a "fluorescence" image is true fluorescence and how much is leaked reflected light. The overlap can be large: with a royal blue LED behind the blue filter, up to 41.0% of the LED's light got through.

Light from a royal blue LED passing through the blue filter: a large overlap. Library of Congress
What they tested
| Material | Result |
|---|---|
| Whatman recycled papers, aged and unaged | the paper's own moderate fluorescence, and unknown sample densities, made it impossible to separate paper from sample |
| Herblock pigments | no fluorescence beyond what leakage could explain |
| Modern pastels | chosen for strong fluorescence |
| The "Ptolemy Map" (folio 96 recto) and colorants like those thought to be used on it | substantial differences between reflectance and fluorescence images in several areas |

The Ptolemy Map, folio 96 recto, under green light (535 nm): (a) no filter, 1.6-second exposure; (b) red filter, 30-second exposure. Library of Congress
In the region labeled "Africae Minoris Pars," the map looks red partly because something in it converts green light into red fluorescence. What that fluorescent material and pigment are has yet to be identified.

A color composite of the region of the Ptolemy Map used to assess fluorescence. Library of Congress
Crossed and parallel polarizers were also used to image Library documents, though polarization signals were largely absent.
Outcomes
- Procedures and protocols for fluorescence studies, and new software modules for analysis.
- Seminar and training materials; 12 preservation professionals were trained in September 2011.
- Plans to add more filter bands for richer materials characterization.
Heritage science adopts new technology slowly, by necessity — it must avoid risk to irreplaceable objects. Folding new capabilities into proven techniques, the team argues, is the best way forward.
Sources
Based on "Materials Characterization Utilizing Advanced Spectral Imaging," Library of Congress, Dr. Fenella G. France, principal investigator, with Meghan Hill and Dr. Bill Christens-Barry (grants MT-2210-09-NC-0 and IA-2210-10-0139), as posted by the National Park Service. Rewritten in hubnx's own words; the technical tables and most figures are summarized rather than reproduced.
Licenza: CC0 1.0 (pubblico dominio) · Tratto da www.nps.gov
1
0
0
0

Commenti






