Hi,

I think a desktop UI for OCRopus would be great.  We're not working on a
desktop app, but we are hoping to be able to develop some web-based apps
specifically for book capture and digital library applications.

In designing a good UI, I'd suggest thinking carefully about uses cases and
scenarios.  What's the most frequent thing people will want to do after
scanning?  What's the quickest and most intuitive way of doing that?
There's some overlap between desktop scanning and book capture, but also
some differences.

I think the most frequent operations are likely going to be:

   - scanning directly from within the OCR application
   - switching between scan and OCR view and/or overlaying them
   - deleting bad scans
   - reordering pages
   - removing noise regions (page borders, stains, etc.)
   - fixing page thresholding
   - marking regions as images or text (if they have been misclassified)
   - adding or removing column separators
   - cutting or joining text lines
   - fixing OCR errors

Another very important operation would be some support for OCR training on
scanned data and/or corrected data.

In terms of platform, I'm not sure C++ is the best way to proceed unless
you're really comfortable with it; something like PyGtk or Mono/Gtk# is
probably faster and easier to get started with and maintain later.

Whatever language you use, I'd recommend designing the user interface in
Glade; that way, it's easy to modify menus, layout, and interactions without
changing the code.  It also helps if you think of the UI not so much as
developing a big application, but instead as developing a bunch of
components that can be quickly plugged together into a UI.  For example, you
might develop a separate page reordering component and an image correction
component that people could reuse independently.  Focusing on such
individual components also makes developing and testing easier.

In any case, those are just suggestions.  What do you think?

Tom

On Sat, Oct 18, 2008 at 11:57, Filip Dominec <[EMAIL PROTECTED]>wrote:

>
> I have created a draft of the possible GUI for Ocropus:
>
> http://ocropus.googlegroups.com/web/ocropus-gui-draft1.png?gsc=l5a2YwsAAAC86osmIgM5905GDO3P-MwZ
>
> I know that the GUI will need a lot of months of work before it will
> be usable. However, in my opinion it is important to develop the
> backend and frontend simultaneously to maintain good interconnection
> between the two applications.
>
> 1) Do you also consider such software useful? (Haven't I missed some
> existing Ocropus GUI project?)
>
> 2) I would start writing this application in C++ with GTK+, to make it
> fast and portable on most platforms.
>
> 3) If you think up any contributions to this idea - and I am sure
> there are many thinks to improve - write it here! The image has been
> made in Inkscape - feel free to edit it and upload your better
> version.
>
> 4) If there was even some C++ programmer to join, i would be
> grateful.
>
> >
>

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"ocropus" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/ocropus?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to