On Thu, Nov 12, 2009 at 1:07 PM, John Smith <[email protected]> wrote: > 2009/11/12 Peter Ross <[email protected]>: >> It's great, but I at least in this location it's out by about 3-5m. > > Considering that's bordering on GPS accuracy it's kind of hard to tell > the actual accuracy that way. > Sure if you only have one GPS track for that location, however I have multiple gps tracks taken on different days over about 2 months, so I'm assuming that using the mid point of all of those traces should be where the track is. If you use josm and download the gps traces for the region I'm talking about you should be able to see what I mean.
> Brendan did some comparisons with the property boundaries in QLD and > came up with an accuracy of 1-4m. > Ben Last from nearmap got in contact with me after I posted the above on the nearmap wikipage and I quote: "We'd like to find out what sort of errors you're seeing. The Melbourne survey wasn't flown with precision GPS due to a fault, but it shoudl still be within 5m or so absolute." After that I did some more accurate measuring and the underlying imagery needed to be moved 6m east and 3m south and then it all lined up extermely well. >> However the actual imagery is fan-bloody-tastic! > > Considering other imagery sources are out 20+ m in places so is the accuracy > :) > It's the problem of having such good imagery, when things are out it's much more noticeable. _______________________________________________ Talk-au mailing list [email protected] http://lists.openstreetmap.org/listinfo/talk-au
