My inclination is..
1. Document accurately the process you have used
2. Note the activity required to increase the accuracy (reefs, bay closure)
3. Upload the borders currently there (things aren't working properly
right now).
4. Work to increase the accuracy.
5. Update the borders accordingly.
Ian.
On 22/09/12 03:24, Michael Krämer wrote:
Hi,
finally I got the process sorted out to generate a border from
natural=coastline, natural=reef and the baseline segments. The
resulting osm-file can be found at [1].
Using the dataset I also updated the picture from the post below to
show the differences [2]. It doesn't look that bad everywhere but
especially along the Queensland coast that's not the only place like
this.
So no my question is how we should move on from here. I see different
options:
- Upload the dataset "as is"
- Try to get it better e.g. by
* manually by adding baseline segments (close bays, low water)
* do a more accurate buffer computation
* trace more reefs from Bing
* ...
- dump the idea and look for a better one
- ...
BTW this time I've taken notes so if anyone is interested I could
share the process I've used. The processing hasn't been too acurate
given the limitations of the data. Especially I used a buffer of 0.2
degrees instead of exactly 12 nautical miles so the distance is always
a bit too short.
Michael
-----
[1] https://dl.dropbox.com/u/3882550/OSM/NationalBorder.osm.bz2
[2] https://dl.dropbox.com/u/3882550/OSM/ComputedCoastline.png
_______________________________________________
Talk-au mailing list
Talk-au@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk-au
