On 16/02/2015 12:53, James Grimster wrote:
> +1 to IIIF image and presentation API.
Definitely. IIIF has been seeing widespread adoption in the library
world, and it's great to see museums getting on board (there was a lot
of talk about IIIF at the EuropeanaTech conference last week).
> Look for the IIPImage / IIIF integration work that Klokan have been doing.
Agreed. IIPImage is a very capable server, and the 'trunk' now has IIIF
support thanks to Klokan.
N.b. however, the dirty secret of IIPImage is that although it is fully
open source, if you want to use JPEG2000-format images as the master
representations, it needs a proprietary piece of software called Kakadu
<http://kakadusoftware.com/> to convert between JPEG2000 and other formats.
So, to use it legally, there only seem to be the following options:
1) Use pyramidal TIFFs instead of JPEG2000 masters (this is what we do
for http://collection.whitney.org )
2) Buy a license from Kakadu and optionally recompile from source (this
is what we do for the Qatar Digital Library http://www.qdl.qa/)
3) Buy pre-compiled binaries from Klokan as they have a license to
redistribute Kakadu binaries (this is what I discovered from Petr
Pridal, MD of Klokan when I spoke to him last week)
There is a fully open source project called OpenJPEG which can also do
this conversion, but it is apparently currently several tens of times
slower than Kakadu so probably not suitable for production use.
> delivery over Wellcome/ British Library universal player
> one challenge East Sussex The Keep archives team and us are working on is how to semi-automate OCR into annotations / manifests without
> getting into seriously sophisticated workflows.
> we'll keep the list posted with outcomes
I would definitely be interested in hearing how you achieve this. It's
unclear to me at the moment how transcriptions with detailed positional
information should be represented using the IIIF Presentation API.
E.g. for the Qatar Digital Library we have OCR transcriptions in ALTO
<http://www.loc.gov/standards/alto/>-format XML that include positional
information. We currently represent this manually online by using some
client-side processing (see e.g. this page
<http://www.qdl.qa/en/archive/81055/vdc_100022698519.0x000010#transcription>,
then open the 'transcription' panel and click 'Apply Page Layout') but
it would be great to do this in a standard format that would also mean
we could overlay this on the source image.
-Tristan.
--
cogapp
Tristan Roddis
Head of Web Development
+44 1273 821 600
www.cogapp.com <http://www.cogapp.com>
*News*
Qatar Museums’ website recently won an international award for Best
Cultural Experience. Find out more here
<http://www.cogapp.com/news/qatar-museums-website-wins-best-cultural-experience-award>.
For more Cogapp news, please follow us on Twitter
<https://twitter.com/cogapp> and LinkedIn
<https://www.linkedin.com/company/cogapp>
****************************************************************
website: http://museumscomputergroup.org.uk/
Twitter: http://www.twitter.com/ukmcg
Facebook: http://www.facebook.com/museumscomputergroup
[un]subscribe: http://museumscomputergroup.org.uk/email-list/
****************************************************************
|