Extraction and Classification of Scriptfeatures on handtorn Paperfragments II

01.01.2010 - 30.06.2010
Assigned research project
The project aims at the preparation of individual paper pieces (fragments of handtorn documents) for a subsequent classification which is needed for the virtual reconstruction of these documents. Furthermore, a number of features should be extracted from the visual information. The paper fragments are available in a scanned version as a color images. In addition, there is a binary mask for the distinction between fragment and background available. The main objective of the six-month project is the classification of the script- and symbol content of the paper fragments in Script, symbol and paper background. Other features to be delivered are rotation of the page, writing direction, font color, Lineanalysis (Lines, squared, without) as well as font (typewriter, manuscript, symbol).

People

Project leader

Project personnel

Institute

Contract/collaboration

  • Fraunhofer-Gesellschaft zur Förderu der angewandten Forschung e.V.

Research focus

  • Media Informatics and Visual Computing: 100%

Keywords

GermanEnglish
KlassifikationClassification
BildverarbeitungImage Processing
SegmentierungSegmentation

Publications