OziExplorer CE and map calibration/precision...

polyfusion

Was wondering if anyone here uses OziExplorer CE with maps they've calibrated themselves?

I calibrated a map using co-ordinates obtained on Google Earth (using the centre of roundabouts/bridges/ponds/etc) but the calibration is a little out, maybe 10-20 meters (looks like I'm driving on the right hand side of the road driving north, and overshooting junctions by about the same amount when driving east for example).

Is this about as good as I'll be able to get it? I converted seconds correctly for Ozi (/60, *100 - was about 500-1000m out before I did this - I was rounding them to the nearest half second!).

And should I be trusting Google Earth for the co-ordinates? I noticed when I overlay roads on the satellite imagery, you can see that they're not exactly superimposed. I used the satellite image for my points; maybe I should use the roads instead? This might account for some of the descrepency I'm seeing. One point I used was one of the roundabouts on the new Grangecastle road, the second one in from the Naas DC, it was obvious at that point.

eg: http://maps.google.com/maps?q=53.307878,+-6.424980&num=1&t=h&vps=1&jsv=178b&sll=37.0625,-95.677068&sspn=23.875,57.630033&ie=UTF8&hl=en&geocode=FeZpLQMdbPad_w&split=0

101sean

Your map datum will be incorrect. I had this trouble but a good read of the FAQs on the ozi explorer site and a mess around with the options sorted it.

For OSI, the map datum should be Ireland 1965 and your GPS datum should be WGS84. I used sat nav mapping for my tie in, Google uses Teleatlas, Navteq based mapping will be more accurate for Ireland. Don't know how well the public satellite imagery is matched.

polyfusion

Thanks for the reply.

Actually, my map datum was correct, turns out the 6 or 7 tenths of a second difference between google maps satellite imagery and the overlaid roads was causing the discrepency. I was initially using the satellite imagery; I switched to the raods, like in the middle of a crossroads for a calibration point. It's spot on now, can even notice the difference in the resulting tracks from driving in either lane of a dual carrageway, going the same direction, so I think it's accurate enough for me now.

101sean

Glad that's sorted, Ozi's website reckons the majority of problems are down to datums but obviously not in this case!

From what I recall from when I was involved with GIS systems years ago, satellite and aerial imagery has to be ortho-corrected, center of the image under the camera will be correct but the sides will be distorted. Suspect Google are using uncorrected images.