Second Step: Page Regions
On one page there are structurally and contentwise different regions. Once the print space has been determined and labelled, the layout is analyzed. Different regions have to be distinguished and labelled as such:
- Text region : TextRegion,
- Illustrations, photos : ImageRegion,
- Book decoration, drawings : GraphicRegion,
- Separator lines, separators : SeparatorRegion,
- Tables : TableRegion,
- Line drawing : LineDrawingRegion,
- Maps : MapRegion
- Mathematical formula : MathsRegion,
- Chemical formula : ChemRegion,
- Musical notes : MusicRegion,
- Advertisement : AdvertRegion und
- Dirt, stains, noise : NoiseRegion
How profoundly a text can be transcribed is determined by the transcription level.
Important: It is not sufficient to identify the entire print space as a TextRegion as
the individually delimitable blocks must be localized. This also makes it possible
to
clearly distinguish between textual and non-textual segments.