Re: [Jocr-devels] Weak database

[Joerg <Joe...@UR...>]

Hi Michael,

You probably dont find databases because the database method is 
in an alpha state.
That means the database option is usable if everything else is failing
but the database can save the day.
Therefore I can not recommend to build a world wide database for gocr.
Nevertheless I am interested on such ideas and can try to put it into
usefull code.

What I can do for you is improve the default algorithms of gocr for your 
documents. Especially if your documents are of good quality I would like 
to have an cut-out of the nonrecognized text to analyze the algorithm
and fix the problem.

Regards,
 Joerg.

On Tue, 7 Jun 2005, michael wrote:

> Hi All,
>  
> I am trying to create a document managment system using ocr and saving it
> into a database, I am having problems with gocr picking up most my letters.
> I have searched the web for other users /db/s this would be the best way to
> build up a database, but this was unsuccessful I found nothing.
>  
> Could users from the mailing list that have got a really good and nice
> working database tar them up and submit them to the forum or put them on a
> site somewhere for the public to use.
>  
> Thanks,
> Michael da Silva Pereira
>  
>