[OSM-dev] TIGER import

James Marca jmarca at translab.its.uci.edu
Fri Jun 8 20:11:27 BST 2007


On Fri, Jun 08, 2007 at 01:30:31PM -0500, Alex Mauer scribeth thusly:
> Dave Hansen wrote:
> > One thing I just noted was that the TIGER data seems to be ~60m off for
> > a small street in my neighborhood.  anybody else see those kinds of
> > errors?
> 
> I often see errors like that in individual streets.  I generally see
> them on newer streets, perhaps because the road was not built as
> originally planned.

We use 250m bound in our map-match code just to be safe---meaning we
select all road segments within 250m of a GPS point as *possible*
matches at the start of our algorithm.  I can't say whether a slightly
smaller bound is more optimal, but 100m definitely missed some
streets.

I've also casually observed that the grid networks in Orange County,
California tend to be more accurate than the "curvy" streets.  The
former are newer than the latter, but the latter are also more
difficult for the US Census to correct.

James




More information about the dev mailing list