I have a 50 page family history book published by a family member in the 1950s.
Looking for inexpensive solutions/solutions for digitizing it that Dopers have used.
I have tried the OCR utility included with current Canon printer/scanners without much luck.
What OS?
I am fairly agnostic but no Linux at the moment. Win7, Win10, MacOS or IOS.
What do you mean by “not much luck”?
can you take the book apart to scan flat pages?
For me, 50 pages is on the outer limit of doable with a medium quality OCR program. There will be lots of errors, but after OCR cleaning them up doesn’t take a lot of time. It gets frustrating, but 50 pages is doable. Of course your milage may vary. I have found that FineReader from ABBYY Production does a good job for medium quality work. As always, you get what you pay for. Prices range from 0 to 1000’s. I don’t recommend 0. Your time is worth more than that.
50 pages? I’d look at one of the book scanning services, rather than buy hardware and fool with it.
For example, these folks would charge you $21.95 to scan it + 0.09/page, for a grand total of $26.45.
Poor OCR
I think that’s the right answer. I think a decent OCR software by itself would cost more than that.
I don’t know what experiences you have with OCR, but maybe you have unrealistic expectations about accuracy? The mentioned Abbyy is very good, but all OCR is GIGO. Unless you have a book scanner or are willing to disassemble the book, there is going to be a problem with the gutter edges of the pages. (I used to have a Plustek OpticBook, but it died years ago.)
But there is another path that you might not have considered–there are lots of OCR apps for Android (and I assume iPhones) that use the phone’s camera to capture the page and then OCR it. Try a few of those (and there are lots more than those listed) but know that nothing will not require lots of hand-editing.
Been there, done that, for almost exactly the same purpose (a book of family history). I just used the OCR software I had and fixed the errors manually… but that sometimes meant completely re-typing a half a page of meaningless grawlix. It probably didn’t help that different portions were written by different people, on different typewriters/computers, and the copy I was working from had been through at least one and probably more generations of photocopying.