There are multiple typos that need correcting but they seem to be generated by the scanner, but not all of them!
I'm confused. Are you saying that the software is decoding / encoding the scanned image into a text format? Text recognition in other words.
I guess you haven't got the original pages. It would be a shame if they were scrapped upon transcribing.
I have many of my father's letters to my mother from WWII, I intend to scan them as jpg's and archive on DVD. Rough edges, mud, tea-stains and all.
This is by the way, but I also have a charming little Birthday Book. Many, many ancestors in there. My mother was one of the last when she signed it at age 12 -- born 100 years ago.
The book is hard-back and of course very precious so I didn't want to stress the 'hinge', yet the pages lie flat enough to get a good focus to the spine.