OpenStreetMap
I have made a patch for inkscape to paste in the openstreetmap way links for photomapping.
"create link to OSM way id in clipboard" is the new right click menu item,
it takes the current osm way id and makes a link like this in svg :
HRef:
http://www.openstreetmap.org/browse/way/667599314
Title:
Openstreemap Way number 667599314
Target : new
Code is here:
https://code.launchpad.net/~jamesmikedupont/inkscape/osm-inkscape
Demo Video:
http://www.youtube.com/watch?v=UApXq5_0yvk
Discussion
Comment from Chaos99 on 8 April 2010 at 05:48
Hi, I'm following your efforts in photo mapping for a while now, but I can't really see what it's all about yet.
Do you have a site where you explain the whole concept or the goal you are aiming at?
For now all I can see is that you link portions of an image to (existing) objects in osm. (By using inkscape and josm). But what's your workflow to ultimatly create map data from the image? Shouldn't it be the other way around? Creating new osm objects for parts of the image?
You see I don't quite understand what you are doing, but I'm very interested to find out.
Comment from h4ck3rm1k3 on 8 April 2010 at 06:25
Well, here is my plan :
1. be able to take the osm id from the clipboard (in josm ctrl c on an object) and create a link to that way.
2. be able to download the geometry of the osm object and add it as attributes of the object.
3. be able to assign geometries to points as well
4. be able to determine the coordinate of new items from the existing geometries (a 3d grid)
5. be able to copy new items from inkscape into josm
The idea is this:
if you know the fixed point of some items, and you know the relative positions of others from the photos, you can add in the new relative items.
If you have a row of houses along a street, then you can just add them in to osm one after another to fill up the street according to the geometries of the photo.
Then if you have them tagged on the photo, it should be possible later to do some type of sanity check on the coordinates, each pixel will have some type of size associated to it, according to the geometry of the map.
You can then pull the information back from osm, into the photo and check them, does the size of the house match the box on the picture in any way?
In any case, this is what I have been experimenting with.
Even if we dont have software to do this, we have brains, and our brains should be able to check the relationships.
You can see this area of the map
http://www.openstreetmap.org/?lat=41.989812&lon=20.705707&zoom=18&layers=B000FTF
All those houses I added in from the photo. and you can see the graveyard and the streets as a point of reference. Even if they are all shifted off in some direction we can fix that when we have more data.
If you add in all the buildings then it is then a matter of verifing them. But 10 buildings that fill up a street will have to have some type of even spacing to them.
mike
Comment from h4ck3rm1k3 on 8 April 2010 at 09:37
Here this changeset here :
http://www.openstreetmap.org/browse/changeset/4362442
was derived from this photo here :
http://www.flickr.com/photos/tonythemisfit/2443746612/