Hello all,

I would like to do some OCR in Linux. I have decided to try GOCR. I
get a result which is not that good. But I am hoping that it could be
bettered by doing some tweaking but I don't know which tweaking !

So far I have done things the following way :

300 dpi scan in grayscale with Sane,
Saved the image in pnm file.
Ran through gocr with no switches.

I also tried with a pbm file that i concerted from the pnm file.

The result was better with the pnm file although not very much better.

Any ideas how I could better my results ?

Thanks in advance,