- From: Raphaël Troncy <Raphael.Troncy@cwi.nl>
- Date: Thu, 01 Mar 2007 13:53:45 +0100
- To: Thomas Franz <franz@uni-koblenz.de>
- CC: Susanne Boll <susanne.boll@informatik.uni-oldenburg.de>, erik.mannens@UGent.be, MMSem-XG Public List <public-xg-mmsem@w3.org>, rik.vandewalle@UGent.be, davy.vandeursen@UGent.be, Richard Arndt <rarndt@uni-koblenz.de>, Michael Hausenblas <michael.hausenblas@joanneum.at>, Christian Halaschek-Wiener <halasche@cs.umd.edu>
Dear Susanne, Erik, all,
Thanks Thomas for the pointers you sent, I really like this work :-)
Erik, I find your plan really good. Do you know how much time you will need to do such a
conversion? What are your concrete plans? Who is involved in the specifications of DIG35
v1.0 and do you plan to contact them to have their support? Have you any ideas of who use
and support DIG35 v1.0 (which applications?). And will it be possible to get a consensus
for such a conversion?
Would you like to fill the section 3.1.4 DIG 35 in
http://www.w3.org/2005/Incubator/mmsem/deliverables/vocabularies.html ?
Susanne,
> > thank you for the MMLab overview. However, I would still like
> > to discuss the question if RDF is reasonable for PhotoMetadata
> > or how we handle metadata that is difficult to be described with
> > RDF, e.g., areas in photos (as a resource in RDF) annotated with
> > a person name.
Hummm, I better understand what your problem is ...
Actually, this problem is already addressed (but not entirely solved) in the "Image
Annotation on the Semantic Web" document,
http://www.w3.org/2005/Incubator/mmsem/XGR-image-annotation/
More precisely, your problem is not in the annotation per se (that you can represent in
RDF or whatever) but on how to identify uniquely a particular region of an image ! The
previous deliverable mentions a number of ad-hoc solutions:
- You can define your boundinx box in SVG ... and then, the RDF annotation is about
this piece of XML that described the boundix box ! This is the solution used by
PhotoStuff, Chris, correct me if I'm wrong ...
- You can defined more precisely the region (not only a rectangle but any shape) in
MPEG-7 ... and again, your RDF annotation will be about this piece of XML that describes
the region !
- You can use a full RDF solution using the MPEG-7 ontology designed by Richard Arnt
and me in K-Space.
- You can imagine other kind of solutions :-)
Personally, I would advocate another ideal solution ! The SpatialURI following the
TemporalURI [1] (soon a RFC) scheme.
A SpatialURI would define how to uniquelly identify, describe and dereference a region of
a still image, the same way that TemporalURI can specify how to describe, serve,
dereference a temporal segment of a video !
Best regards.
Raphaël
[1] S. Pfeiffer and C. Parker and A. Pang. Specifying time intervals in URI queries and
fragments of time-based Web resources, Network Working Group, Internet-Draft, 2003,
http://www.annodex.net/TR/draft-pfeiffer-temporal-fragments-02.html
--
Raphaël Troncy
CWI (Centre for Mathematics and Computer Science),
Kruislaan 413, 1098 SJ Amsterdam, The Netherlands
e-mail: raphael.troncy@cwi.nl & raphael.troncy@gmail.com
Tel: +31 (0)20 - 592 4093
Fax: +31 (0)20 - 592 4312
Web: http://www.cwi.nl/~troncy/
Received on Thursday, 1 March 2007 13:01:40 UTC