[Simh] DIY microfiche scanner?

[Mattis Lind]

2018-02-23 15:11 GMT+01:00 Bob Eager <rde at tavi.co.uk>:

> On Fri, 23 Feb 2018 08:05:06 -0500
> "Tim Stark" <fsword007 at gmail.com> wrote:
>
> > I recently noticed that you recently talked about DIY scanning for
> > documents.   Does anyone have DIY techniques about microfiche
> > scanning?
>
> I'd love to know. At the moment I have a large microfiche reader which
> is taking up far too much space. Got that from eBay.
>


I am also interested in this topic. I was recently given a huge number of
microfiche by a nice gentleman. Thanks Mike! DEC Tech manuals, Diag
listings, IPBs etc etc. In order 5000-10000 fiches in three blue steel
boxes.

I had an idea of making a catalogue of available fiches. I started of
taking a picture of them just for the purpose of be able to write down the
information later on (Just the tech manuals are approx 1300 fiches:
https://www.dropbox.com/sh/8cgznonlixavlsx/AABQI2-sQuxcqBO1dHcAFMUga?dl=0).
Then this information can be shared and those documents that are not
available can be scanned at a later stage. I simply don't see any point in
scanning them all (if I had a scanner that is) since many documents are
already online.

But the sheer number of fiche already tell me this is not feasible to do
this manually. I need some kind of automatic method of doing this.

I was thinking some kind of pipeline of steps that takes the image and
converts it to a database entry or spreadsheet row. Identify fiche outline,
Straighten it up. Identify text locations. Do OCR. Identify type of text
based on text contents etc.

https://drive.google.com/open?id=1c_8TFNDkPd8poigdbuJiohDod5Z08ifpD9lqXNRPMeA

I have recognized a couple of different fonts used and the font size varies
slightly. The positions are relatively fixed. The position of the date and
Copyright year and format varies a bit though.

Anyone did something similar? Ideas? Useful software to use?

/Mattis


> _______________________________________________
> Simh mailing list
> Simh at trailing-edge.com
> http://mailman.trailing-edge.com/mailman/listinfo/simh
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20180224/c6679ac7/attachment.html>