[MM] NASA use case solution via PhotoStuff

Hi Nikolaos,
   Jacco has asked me to sketch a brief overview of a potential  
solution for the NASA use case using PhotoStuff. Maybe this will help  
in your tool evalution, etc...

Anyway as I stated in my previous email (which included a high level  
solution to the NASA use case), we can use PhotoStuff to provide the  
RDF/XML annotations of image content. Essentially, users can load  
images and ontologies into the tool and markup the images (and/or its  
regions) as they wish. In a previous email, I referred you to the  
following sample image residing at http://grin.hq.nasa.gov/IMAGES/ 
SMALL/GPN-2000-001171.jpg. Lets assume that we would want to assert  
the the image depicts the Apollo 7 launch. Also assume that want to  
say that the Apollo 7 Saturn IB space vehicle is depicted in a  
rectangular region around the rocket. We also want to assert that the  
image was taken on 11/11/1968 and was created by NASA; we can then  
state that the image was taken at Kennedy Space Center in Florida.  
Lastly, we can assert that the image is 451x640 (width x height in  
pixels). This type of image annotation can be provided by PhotoStuff.  
The image can be loaded into the tool, as well as any number of  
related ontologies. Let assume that we will use the following space  
related ontologies:
     Shuttle related
         http://semspace.mindswap.org/2004/ontologies/ShuttleMission- 
ont.owl

     Space Vehicle System related
         http://semspace.mindswap.org/2004/ontologies/System-ont.owl


These ontologies and the image into PhotoStuff, by simply typing the  
URLs into the address bar. Then the user proceed to markup the image.  
I have performed the annotations required above. I will avoid  
detailing the steps necessary for accomplishing this...refer to [1]  
for a user guide.

The resulting RDF/XML annotations are attached to this document.  
Image depictions are asserted via a depiction property (a sub- 
property of foaf:depiction) defined in the MINDSWAP Digital Media  
ontology [2]. Thus image regions can be semantically linked to  
existing or newly created instances defined on the Web. The image  
regions are defined via a property named svgOutline. Essentially SVG  
outlines (SVG XML literals) of the regions are specified using this  
property. Thus PhotoStuff (and any advanced metadata browsing  
environments intended to be used to browse the annotations, etc.)  
displays the image regions based on the SVG outlines defined using  
this property. Anyway, the rest is fairly straight forward. Please  
let me know if you have any further questions.

The resulting metadata produced from such annotation sessions can  
then be shared, indexed, and used by advanced annotation-based  
browsing (and searchable) environments.

I think thats about it. Let me know if you need any clarifications, etc.



Cheers,
Chris


References

[1] PhotoStuff User Guide. Available at: http://www.mindswap.org/2003/ 
PhotoStuff/UserGuide.pdf
[2] MINDSWAP Digital Media Ontology. Available at: http:// 
www.mindswap.org/2005/owl/digital-media




-- 
Christian Halaschek-Wiener
PhD Student, Dept. of Computer Science
GRA, MINDSWAP Research Group,
University of Maryland, College Park
Web page: http://www.mindswap.org/~chris