At Wednesday 20/02/2002, you wrote:
>Apologies for cross posting.
>
>Could colleagues share their experience of successful use of OCR
>software? We have A5 index cards, typed in assorted fonts, 4 or 5
>columns across, with a few handwritten markings (numbering and ticks in
>pencil or biro). I'm looking to migrate the information to a table in
>Word whenever we have to update a card. We update modest quantities,
>maybe 40 a month. It seems to make sense to scan the old card
>first. Many thanks.
>
>Dr Wendy Sudbury
>Director
>Church of England Record Centre
It is highly unlikely that you will find any OCR package which would give
you acceptable levels of accuracy and thus not prove more expensive than
just getting the records rekeyed either yourself, by volunteers or rekeyed
by a bureau.
The reason is that the total process of OCR has to be accounted for and
will probably work out too costly. First you have to scan the index cards
to get a electronic version for OCR to work from. Then the OCR package has
to successfully recognize every character on the index cards - for
handwriting this is not currently practicable and also likely to introduce
misread characters into the remainder of the OCR. Finally you then have to
find a way to structure the output so that it fits neatly into your Word
table and it is highly unlikely that the index cards are laid out in a
totally uniform way that will allow that structuring to be
automated. After all this I would suggest you would have an accuracy rate
of about 60-70% and thus have to check and correct every record.
Although it may seem the long way round and we would all like total
automation the reality is that for highly variable originals, with
handwriting and variable text and structure to be converted to a structured
accurate digital form requires human intervention. Rekeying will work out
faster in the end.
Best regards,
Simon
============================================================
Simon Tanner
Senior Digitisation Consultant (HEDS)
Higher Education Digitisation Service
University of Hertfordshire
Phone: +44 (0) 1707 286078
Fax: +44 (0) 1707 286079
Web: http://heds.herts.ac.uk
METAe Project: http://heds.herts.ac.uk/projects/projectsE.html
|