Magdalena

Hi,

I am a PM for the Core OCR team at Microsoft Development Center Serbia,
developing an engine for optical character recognition. We are in process of
developing automated tool for simulation of the OCR process. The simulator
generates both: bmp images representing pages of the scanned document (input
of the OCR engine) and xml that corresponds to them (output of the OCR
engine). At this point we are able to generate images with single background
color and single text color. The idea is that we expand the simulator by
adding post processing on the images (degrading the images), so they look
more realistic (e.g. old book pages, newspaper pages, poorly scanned docs).
We were thinking of using some existing MS tool (Expression Design tool) for
applying complex effects on bitmaps. The information and knowledge, you can
hopefully provide us with, would be used for generating referent data for
measuring accuracy of our engine.

We are especially interested in the following:
o Is it possible to perform batch processing on set of images?
o Can we programmatically access some of your live effects ("filters")?
o Is it possible to integrate some of your filter libraries into our tool?

We appreciate any help you can give us. Please let us know if you require
any additional information from me.

Many Thanks,
Magdalena Cocovic

Annie

If you send me your email address I will forward it on to a PM on the
Expression Design team. My email address is:

design at studioe3 dot com.

--
Annie
design(at)studioe3(dot)com
This posting is provided "AS IS" with no warranties, and confers no
rights.



"Magdalena" <Magdalena@discussions.microsoft.com> wrote in message
news:FE4A2754-7AD8-4DB5-8FF3-6A1B1CBBA4BE@microsoft.com...
> Hi,
>
> I am a PM for the Core OCR team at Microsoft Development Center Serbia,
> developing an engine for optical character recognition. We are in process
> of
> developing automated tool for simulation of the OCR process. The simulator
> generates both: bmp images representing pages of the scanned document
> (input
> of the OCR engine) and xml that corresponds to them (output of the OCR
> engine). At this point we are able to generate images with single
> background
> color and single text color. The idea is that we expand the simulator by
> adding post processing on the images (degrading the images), so they look
> more realistic (e.g. old book pages, newspaper pages, poorly scanned
> docs).
> We were thinking of using some existing MS tool (Expression Design tool)
> for
> applying complex effects on bitmaps. The information and knowledge, you
> can
> hopefully provide us with, would be used for generating referent data for
> measuring accuracy of our engine.
>
> We are especially interested in the following:
> o Is it possible to perform batch processing on set of images?
> o Can we programmatically access some of your live effects ("filters")?
> o Is it possible to integrate some of your filter libraries into our tool?
>
> We appreciate any help you can give us. Please let us know if you require
> any additional information from me.
>
> Many Thanks,
> Magdalena Cocovic
>