@device(postscript)
@libraryfile(Mathematics10)
@libraryfile(Accents)
@style(fontfamily=timesroman,fontscale=11)
@pagefooting(immediate, left "@c<Technical Report 1995>",
center "@c<Computer Science>",
right "@c<Carnegie Mellon>")
@heading(Top Ten lessons Learned in Automated Cartography)
@heading(CMU-CS-96-110)
@center(@b(David M. McKeown, Jr.))
@center(January 1996@foot<This paper will appear in the @i[Proceedings of the 1994 ARPA Image Understanding Workshop], February 12-15, 1996, Palm Springs CA.>)
@center(FTP: Unavailable)
@blankspace(1)
@begin(text)

Over the last 16 years researchers at the Digital Mapping Laboratory
have worked on a number of topics in automated cartographic feature
extraction.  This research has covered several problem domains: automatic
building extraction, road network extraction, stereo matching and
analysis, digital terrain modeling, analysis of multispectral imagery,
and the development of knowledge-based systems for scene
interpretation.  In this brief talk, I'll share a very personal view
of what we've learned over the years, why progress has been so slow,
and where the action is for the future.

@blankspace(2line)
@begin(transparent,size=10)
@b(Keywords:@ )@c<Automated cartography, lessons learned, computer vision>
@end(transparent)
@blankspace(1line)
@end(text)
@flushright(@b[(4 pages)])