Digital Atheneum Project University of Kentucky

1
Digital Atheneum Project University of Kentucky

• New Approaches to Restoring, Editing and
  Searching humanities collections.
• by
• Arpita Goenka

2
Principal investigators
Brent Seales
James Griffioen
Kevin Kernian
3
Introduction

• It is developing new methods to restore and make
  accessible previously lost writings
• Focuses on recovering the manuscripts of the
  famous Cottonian Library Collection in the
  British Library.
• Many of the manuscripts badly damaged by 1731
  fire and further deteriorated due to neglect,
  misuse and other conservation methods.
• Even with the best digitization techniques the
  remaining of the text is difficult or impossible
  to decipher.
• With damaged manuscripts the search engine must
  be devised to accommodate partial and distorted
  forms.

4
Project focus

• New digitization techniques to illuminate all
  possible information from the original
  manuscript.
• Restoration algorithms to attempt to fill in the
  most likely missing information
• Complex, data-specific, content search techniques
  to identify the imperfect representations found
  in severely damaged manuscripts.

5
Using Illumination to Enhance Digitization

• The images must be obtained using extremely
  high-resolution cameras
• Many parts of the damaged documents are invisible
  without special lighting techniques
• The manuscripts are rarely completely flat or
  planar.

6
Illumination Techniqueusing ultraviolet lighting
to reveal formerly invisible text
7
Limitation of 2-D imagery

• When imaging a non-planar object-it can appear
  warped or crinkled.
• Difficult to disambiguate if warping is part of
  original or an artifact of the objects shape.
• Provides insufficient look and feel
• Items such as wax seals, coins, etc have inherent
  3-D shape

8
3-D Acquisition setup

• Introduce a light projector along with the camera
  to capture 3-D models.
• Projector and camera are used to triangulate 3D
  points on the artifacts surface

9
Depth Information

• Depth information may also help solve the problem
  of accurately reuniting physically separate
  fragments

10
Searching Images with Computational Methods

• Creating document specific image processing
  algorithms that can locate, identify and classify
  individual letterforms.
• By analyzing several letterforms, computer models
  are built that can perform probabilistic pattern
  matching of damaged letterforms.
• Transcriptions aid significantly in searching by
  narrowing the search space and assisting an
  editor who is struggling to decipher a charred
  leaf.

11
Editing and Annotating the Damaged manuscripts

• Encoding the transcripts and edited texts in SGML
  to facilitate comprehensive searches,and are
  converting both to HTML or XML so they can be
  displayed by Internet browsers.
• Developing a generic toolkit to assist other
  editors (like scholars in humanities) in
  assembling complex editions from high-resolution
  digital manuscript data.
• An editor can then collect and create the
  components of an electronic edition for any work
  and use the generic toolkit to fashion a
  sophisticated interface for an electronic display
  of the edition.

12
Editors Tool

• It must be Flexible encompass many different
  collections
• Usable support non-expert computer users
• Technically sophisticated incorporate new
  technical solutions
• Functions
• include Registration, Mosaicing,Textual
  correspondence, Glossary contsruction

13
Refernces

• "The Digital Atheneum - Restoring Damaged
  Manuscripts." RLG DigiNews 36
• "The Digital Atheneum New Technologies for
  Restoring and Preserving Old Documents."
  Computers in Libraries 202 (February 2000),
  26-30.
• "The Digital Atheneum New Approaches for
  Preserving, Restoring and Analyzing Damaged
  Manuscripts." Proceedings of the First
  ACM/IEEE-CS Joint Conference on Digital
  Libraries, 2001. (NY ACM Press, 2001), 437-443.
• http//www.dli2.nsf.gov/cornellworkshop/