wiki-archive/twiki/data/Image/ImageTopicsFromBratislava.t...

214 lines
6.8 KiB
Plaintext

head 1.4;
access;
symbols;
locks; strict;
comment @# @;
expand @o@;
1.4
date 2007.09.21.13.13.00; author BobMorris; state Exp;
branches;
next 1.3;
1.3
date 2007.09.21.10.30.26; author BobMorris; state Exp;
branches;
next 1.2;
1.2
date 2007.09.20.13.32.30; author BobMorris; state Exp;
branches;
next 1.1;
1.1
date 2007.09.20.08.37.09; author BobMorris; state Exp;
branches;
next ;
desc
@none
@
1.4
log
@none
@
text
@%META:TOPICINFO{author="BobMorris" date="1190380380" format="1.1" version="1.4"}%
%META:TOPICPARENT{name="WebHome"}%
These are random notes from two informal meetings in Bratislava at the 2007 TDWG meeting. Main.BobMorris volunteers to begin to sort them out and turn them into proposed tasks towards TDWG process documents for the formation of a charter for an Image Interest Group and Task Groups
There is an RDF representation of EXIF, but EXIF like Z39.87 it is mostly technical metadata; DIG35 supports property lists.
Revisit Adobe XMP as standard for embeding metadata in several formats
Lunch meeting September 21: Annette Olson, Greg Riccardi, Fred Ronquist, Bob Morris identified these concerns each of could be turned into one or more tasks
1. Advise what should be expected of LSID resolvers on image stores.
2. Locate and examine existing best practices manuals about image management to identify which, if any can be endorsed by TDWG
3. Identify what are the concerns of image providers (e.g. IPR, obligations to contributors, obligations _of_ contributors, terms of use...)
4. Standards for transport, especially mechanisms by which the content metadata and image can be robustly associated.
5. Identify other organizations with image management concerns
6. Write and submit TDWG process docuements
From Annette's memory of the meeting:
1. Look at metadata fields
2. LSIDs - one issue is multiple versions, such as sizes, digitally altered
for content, and potentially (or not) formats. Also making sure LSID
documentation incls stuff about images.
3. Look at transfer protocols - incl. Whether image travels within metadata
or metadata within image, whether some protocols go with, and then where do
LSIDs go.
4. Look at copyright and liability issues, incl what happens if users
download and use images inappropriately, and if image libraries make a
mistake during image
5. Look at existing image standards, such as color management, scale...
---
Meeting Sept 22: Gregor Hagedorn, Annette Olsen, Bob Morris
Notes by Gregor.
Use case -
NBII A. Olson
Integration various libraries (BHL, Morphobank, EFG - 60000 in Minalto Gallery, NBII xx in own szstem
KezToNature in DiversitzResources,
Harvesting topics debranding (images are highlighted were the come from), Images coming trhough from outside would have to be branded differentlz than local ones: this image is from - provided through
Which level of metadata is available
Openaccess licences require attribution
Perhaps image and metadata need to be attributed separatelz.
Different cases, all resolutions, or only low res resolutions
Harvesting images
Chris Freeland (BHL) suggested to use [[http://www.openarchives.org/OAI/openarchivesprotocol.html][OAI-PMH]]
BHL at the moment would only be provider)
Bugwood/forestryimages 35000 images of pests online, more coming. Highly desirable if integrator system can provide its own thumbnails. Details of harvesting not available, annette will find out
Annette is revising NBII recommendations on metadata
Gregor - on their site, often,
publish/unpublished
filtering out unpublished items when harvesting - that is the responsibility of the contributor
MorphoBank has this, as does Gregor's Diversity Resources.
Annette can send in base of image galleries document that lists the galleries, # of images, capabilities. will attach, then
could be a task for a group.
issue - what happens if a contributor wants to withdraw images....
harvesting - set up refresh the collection for updates...
how do you handle harvesting from say, the Smithsonian where changes in a collection are made.
can removal of an image be automatic, possibly reharvest the entire collection instead.
the harvesting process needs to be able to know that images need to be removed...
deletion of metadata information and deletion of entire entities...
deletion of target without deletion of information, so can replace....
Gregor uses LinkChecking...
if link not found after protocol, then becomes hidden to public
Annette: we are revising our database to not make images and metadata directly tied to the thumbnail, but to another id.
looking for something when harvesting that flags that a record is changed..., if the contributor can handle it.
what is that status of NBII giving out its interface?
NBII is for open source. definitely the tools could probably be made available. Annette needs to ask about the interface.
but would need resources to package and put out there as the interface is not a formal packaged version, but straight coding.
are there better tools out there anyway? Bob: the image management tools that are public available are quite idiosyncratic.
Annette: we are writing ours to Java this fall because we want new tools that we can't do in php.
Bob would recommend we keep it in php. most widely the most widely open-sourced image gallery is Menalto Gallery in php. our tools could contribute
jsp version that Gregor worked with (developed by...) keeps crashing., problems with tools, having problems finding problem.
for mycology and pathology. Gregor's installation currently holds scans of images from institutional research, out of copyright images from books, images for identification/keys, line drawings, will be part of Quito Nature, 100,000 images of plants, lichens, NHL will be there, more...Estonian university.
---
-- Main.BobMorris - 20 Sep 2007
@
1.3
log
@none
@
text
@d1 1
a1 1
%META:TOPICINFO{author="BobMorris" date="1190370626" format="1.1" reprev="1.3" version="1.3"}%
d3 2
d9 1
a9 1
Lunch meeting: Annette Olson, Greg Riccardi, Fred Ronquist, Bob Morris identified these concerns each of could be turned into one or more tasks
d23 16
d41 4
d68 1
a68 1
Annette is revising recommendations on metadata
d105 1
a105 1
Bob would recommend we keep it in php. most widely the most widely open-sourced image gallery is Renalto gallery in php. our tools could contribute
d111 3
d115 1
@
1.2
log
@none
@
text
@d1 1
a1 1
%META:TOPICINFO{author="BobMorris" date="1190295150" format="1.1" reprev="1.2" version="1.2"}%
d22 3
d26 62
@
1.1
log
@none
@
text
@d1 1
a1 1
%META:TOPICINFO{author="BobMorris" date="1190277429" format="1.1" version="1.1"}%
d3 7
d11 1
d13 7
a19 1
There is an RDF representation of EXIF, but EXIF like Z39.87 it is mostly technical metadata; DIG35 supports property lists.
a20 1
Revisit Adobe XMP as standard for embeding metadata in several formats
d24 1
a24 1
-- Main.BobMorris - 20 Sep 2007@