Robert Sachunsky
Robert Sachunsky
> According to our [glossary](https://ocr-d.github.io/glossary), `region` should be `block`. Agreed! > How about two parameters `input-level` (IL) and `output-level` (OL)? I'm not yet convinced we need both. From the looks...
Another problem with `input-level` is that there are often implicit dependencies between input and output level, so an additional check would have to plausibilize the choices for every processor, raising...
I cannot follow your last comment. Above, I have argued that IL introduces unnecessary ambiguity and necessitates additional checks (as a conceptual and implementatory burden). You answer (in favour of...
> * in the METS specs, new `fileGrp/@USE` name recommendations `OCR-D-IMG-NORM` and `OCR-D-IMG-RAWDEN` should now read: `OCR-D-PRE-NORM` and `OCR-D-PRE-RAWDEN` > * in the PAGE specs, new `AlternativeImage/@comments` classes `normalized` and...
> Then again, our [glossary currently defines despeckling as](https://ocr-d.de/en/spec/glossary#despeckling) > > > Remove artifacts such as smudges, ink blots, underlinings etc. from an image. Typically applied to remove “salt-and-pepper” noise...
> But then the processors should be named accordingly. I would find it quite confusing to use a processor called `denoising` in a workflow step called `despeckling`. Absolutely. Since `despeckling`...
...but also _extend_ it where necessary, e.g. #159, or for language and script identification, or special region detection (only separator lines or tables or stamps or handwriting ...), or pure...
> difference between `layout/analysis` and I always understood that as in logical document layout analysis, not optical page layout analysis. > use them to have an additional means to find...
That's because there are no good table processors in OCR-D yet. But you'd also have to include the existing ones in your workflow in the first place! Here's my take...
> This scan has different skew angles (at top & bottom); perhaps a 3d deskew could help. @jbarth-ubhd, by `3d deskew` you mean dewarping? How did you get this result?...