Fwd: Possible GSoC project idea

[Mukund Raghothaman <muxdevil m gmail com>]

Forwarded to this list for RFC. The idea was liked by folks on
gnome-soc-list, but they asked me to check with the evince team. While
one of Etienne and Joaquim would probably agree to be my mentor, it
would be good to have some support from evince as well.

---------- Forwarded message ----------
From: Mukund Raghothaman <muxdevil m gmail com>
Date: Fri, Mar 19, 2010 at 1:58 AM
Subject: Possible GSoC project idea
To: gnome-soc-list gnome org


Hi,

I'm a student aiming to be one of the lucky few selected by GNOME
under SoC this summer. I've browsed through
http://live.gnome.org/SummerOfCode2010/Ideas, but wish to work on an
idea of my own.

I wish to develop GNOME's support for OCR - I commonly read scanned
documents in evince, and find it a pain to search for text without
Ctrl-F. I would develop a common OCR framework, which together with
plugins for specific applications like evince, eog and GNOME Scan, would
provide a common OCR interface, much like the current print window, or
the scan window provided by GNOME Scan. Underneath the hood, there
would be a pluggable interface that worked with multiple OCR libraries
like Tesseract and Ocropus. As Etienne put it, I'd probably want to
make an API out of what OCRFeeder does, and allow different
applications to use that API.

At a more ambitious level, I want to provide handwriting recognition
support also - a GTK widget that would allow handwriting recognition,
and a similar tool for offline handwriting recognition also. This
would be similar to the tools provided in Windows, and would be of
major help to those using tablet PCs. If this were done, I see people
using Xournal and OpenOffice.org with the same enthusiasm as Office
OneNote. Again, as Etienne says, handwriting is along a different
track, and it would probably work well as a post-SoC project. But
still, my project idea is along a standard document-recognition
interface in GNOME.

I've already contacted Etienne, from GNOME Scan, and Joaquim Rocha,
from OCR Feeder. They asked me to present this idea on this list, to
see what everybody had to say.

Up up? or Down down?

Thanks,
Mukund