OT? Help: Machine manual scan conversion ( OCR and PDF )

Hi,

A while ago , I scanned a Hardinge manual and posted here, some kind fellow took the enormous file I produced and reduced to using a proper Adobe PDF generator with OCR.

I wonder if someone may be able to help again?

I've just acquired and scanned a manual for my Haighton HSU1 milling machine (a rare-as-hens-teeth British copy of the Hardinge UM) and uploaded it to me google drive :

Haighton HSU1 Operators Manual.pdf - Google Drive

Thanks in advance,

Bill

Do you have this on your hard drive? If so download Cutewriter PDf and you can save it yourself as a PDF.

It's less than 1/10th the size when OCR'd. I tried attaching the OCR'd pdf file, but at 460K it exceeds the size limit for this type of file.

RLamparter ,


Thanks. I'll send you an email address via PM

Bill

Thanks Bob.

lovely job

RLamparter, so what did you use for the OCR and prep? Real Adobe PDF authoring software, or some other tool?

I use FineReader 8.0. There are newer versions of FineReader available, but the version I have is pretty accurate on recognition. Where you get into lots of work is when you OCR old publications where the text is blurry or pages with really tiny fonts. Fractions often take some editing.

Thanks, I'll check it out.
How does it do on charts and tables? Something like a feed and speed chart for three or four different cutters, for example. 4-5 columns and 6-8 rows of variable 'height'.

FineReader can do charts, but a lot depends on the font size and the presence of fractions. At least with my older version, charts often need manual adjustment. It's some times easier to frame the chart as a picture and then frame each bit of text with a text box to get an accurate OCR of a chart. If the chart has similar sized boxes such that the rows and columns are uniform, FineReader usually recognizes things correctly automatically without the manipulations I just mentioned. I see it's up to version 14 now. I had tried a demo of a newer version - 12 or 14, I don't remember. There was a learning curve for the newer version, and I don't recall that it was a whole lot more accurate than version 8, so for me it wasn't worth the cost to upgrade. The emphasis on newer versions seems to be on features that would appeal to corporate users and law offices. If you demo an OCR program, make sure you check it on charts with different sized boxes, and on text with lots of fractions. Pretty easy to find both of those in old machine manuals.