[Metadatalibrarians] FGDC metadata and aerial photographs

[Kristin E. Martin]

At UIC, we are planning to embark on a scanning project involving aerial 
photographs of the Chicagoland area, from 1929-present day, about 80,000 
images in all. The metadata that we are starting with for each 
photograph is at most:

Date
Scale
Project #
numeric reference to an index map
Company who took the shot

An engineering firm is going to help us georeference the materials by 
linking the photographs to an index map and providing latitude and 
longitude information.  Once we get them georeferenced, we should also 
have either the bounding coordinates, or a center coordinate point on 
the map, from which we can extrapolate the bounding coordinates and any 
other map project systems we might want (e.g., UTM).  From there we hope 
to use the location information to link up to information about what 
was/is in that location: cities, natural features, buildings, etc. 

The images and their metadata will probably reside in CONTENTdm, and 
we'll try to build an interface to CONTENTdm that is map based to allow 
users to browse spatially.  There might be a possibility to develop a 
more sophisticated spatially-oriented interface, but I suspect that will 
be farther down the road.

I'm wondering about how to do the metadata.  The maps/data librarian has 
suggested the FGDC Content Standard for Digital Geospatial Metadata, 
which I haven't worked with before. From looking at it, I wonder if that 
is too complex to attempt for each individual image.  It also seems like 
a lot of the elements would repeat themselves, because they belong at a 
higher level than the individual image, but there doesn't seem to be a 
way to nest multiple images (where the identification element would be 
different) without having to repeat the other sections.  The levels for 
FGDC seem more appropriate at an aggregate level of the photographs, 
perhaps by year, but then we wouldn't be able to provide the granularity 
of what is in an individual shot.  Additionally, it seems like any 
additional information we add to the metadata, such as buildings, 
natural features, cities, would end up being grouped together in theme 
keywords and place keywords.

I'm wondering if there are other examples out there of aerial photograph 
collections and how you have done the metadata.  Would it be better to 
attempt to the metadata in FGDC and then map selected fields into an 
extended DC format for loading in CONTENTdm, or simply develop metadata 
directly in an extended Dublin Core (better to work with CONTENTdm) and 
worry about attempting FGDC later should we so desire?  Other options?

Thanks,

Kristin

-- 
Kristin E. Martin
Metadata Librarian
Catalog Department (MC 234)
University Library
2-390 Richard J. Daley Library
801 S. Morgan
Chicago, IL  60607
312-413-5052
312-413-0424 (Fax)
kmarti at uic.edu