Re: [Bug-ocrad] Not recognizing obvious text

[Antonio Diaz Diaz]

Tony Maro wrote:
> There's not a way to limit the area of
> the page you're doing OCR on is there?  Like a zone ocr?

Ocrad is able to do layout analysis and process only one of the 
resulting blocks. But this is likely to be slower and less reliable than 
cropping a part of the page as you are doing now.

I have in the to do list for ocrad a "crop" option, and I am in fact so 
impressed with what you are doing that I think I am going to implement 
it in the next version (0.15).

The syntax could be like this:
`ocrad file.pbm --crop left,top,right,bottom'

and the meaning of left, top, right, bottom could be:
between 0.0 and 1.0, a fraction of the whole page
greater than 1, a coordinate.

So `ocrad file.pbm --crop 0.0,0.0,0.5,0.5' would process the upper left 
quadrant, and `ocrad file.pbm --crop 0,0,500,500' would process the 
upper left 500x500 pixel square.


> Bet you guys never thought ocrad would be used for that, eh? ;-)

Really no. :-)


> So, anyone have an idea that might speed up the process?

I think the proposed "crop" option would speed up the process a lot, 
because ocrad could do the cropping first, then the rotation. All 
without creating intermediate files.


> Yes, you read that right.  I'll be processing as much as
> 150,000 pages per day on one server, and am designing this process so it
> could be clustered to handle more.

Some day you have to tell me who are you working for. ;-)


Regards,
Antonio.