dott. Piergiorgio,
I've downloaded the above PDF, and tried a few pages with KADMOS and tesseract. It looks
me like tesseract does a better job on the PDF pages that are converted to TIF's.
$ pdftk amicroprocessori1094517862.pdf burst
convert the PDF pages to *.tif so they can be OCR'd
$ convert -density 400 pg_0014.pdf -background white -despeckle -depth 8 pg_0014.tif
$ tesseract pg_0014.tif p014 -psm 6
Have a look here.
https://github.com/ldkraemer/Eubank-s-EBASIC-source
If no one else works on it, I'll do a few pages each day this winter (November thru March 2022) as I can.
Larry
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)