[OSM-talk] source=(survey, yahoo, gps...)

[John Smith]

2009/9/9 Mike N. <niceman at att.net>:
> There are some odd cases where even GPS averaging is not accurate enough to
> 'micro adjust' a way by 1-3 meters.   One of which is where the sensor is
> inside a vehicle (not rooftop mounted) on some curves.     I observe this by
> comparing multiple traces from multiple units and  the fact that they do not
> converge on a very good aerial image - even though the entry and exit traces
> match very well.

Assuming the aerial image has been aligned with known points to begin with :)