Dataset

TriGraphSlant - benchmark set for writer identification - writers were asked to write in unnatural slant

Brink, A. (Creator), van Batenburg, R. (Creator), Niels, R. (Creator), van den Heuvel, C. E. (Creator), University of Groningen, 18-Mar-2011

Dataset

  • Axel Brink (Creator)
  • Roeland van Batenburg (Creator)
  • Ralph Niels (Creator)
  • C.E. van den Heuvel (Creator)
  • Lambert Schomaker (Supervisor)
  • Donders Institute for Brain Cognition and Behavior
  • Nederlands Forensisch Instituut

Description

This is the TrigraphSlant (Img version) Distribution, release 18/3/2011

This distribution contains 188 images of scanned handwritten text, scanned at resolution 300dpi Canon LiDE 25, grey scale, by 47 Dutch writers, four pages per writer, from four writing conditions, one condition per page. The conditions are:
1. [AN] Copy text A in your natural handwriting.
2. [BN] Copy text B in your natural handwriting.
3. [BL] Copy text B and slant your handwriting to the left as much as possible.
4. [BR] Copy text B and slant your handwriting to the right as much as possible.

The codes AN, BN, BL and BR refer to subsets into which the collected pages of the writers were subdivided. AN represents a collection of
authentic documents; BN, BL and BR can be seen as collections of questioned documents. To avoid structural effects of fatigue, the order of item 3 and 4 was randomized at each collection: half of the subjects wrote the BR page before the BL page. The data were collected at three sites, in three cities: The Hague: NFI (N...), Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen (D...)
and the Artificial Intelligence Dept. of University of Groningen (R...)

Copyright The International Unipen Foundation, 2010, All rights reserved
Date made available18-Mar-2011
PublisherUniversity of Groningen
Date of data production18-Mar-2018
Geographical coverageThe Netherlands
Access to the dataset Open
Contact researchdata@rug.nl

    Keywords on Datasets

  • Handwriting biometrics, Writer verification, Slant, Disguise, Statistical pattern recognition

ID: 64099286