[Openstreetmap-dev] Re: [Openstreetmap] Editor question

[Lars Aronsson]

Tom Carden wrote:

> For an extreme example of what the map would look like under 
> this kind of scheme, take a look at the poster Steve and I made 
> from raw GPS data. Erroneous/extreme jumps aside, which of the 
> 203 journeys down Oxford Street is best, and how do you choose 
> automatically?

Again, Oxford Street is one case, and a country road is a 
different one.  Five tracks along the same road visualized as 
rubber bands could be tied together and would mark the average 
route.  Hopefully the drawn map reflects that average route.

Now a sixth tracklog is added, apparently going down the same road 
but hinting that the drawn road should be moved a little to the 
right or left, or pulling the previous average to a new position.  
Should our drawn streets maintain a weight, the number of tracks 
weighed into the current average, so that we could compute the 
correct pull from any newly added track?

Car navigators don't have better GPS receivers, but they have 
street maps and adjust their position to the nearest street, 
because it is unlikely that someone drives between streets. With 
the current Open Street Map we too have this street map, to which 
we can compare (and adjust?) our most recent measurements.

Accept the first track as a drawn road (as JOSM does), and set its 
weight to 1 (drawn road based on 1 track).  When the second track 
is added, it pulls this drawn road (line segment) to a new 
position and sets the weight to 2.  When the new track deviates 
more than, say, 20 meters from the old one, it is a 
branch/intersection and heads off in another direction.


-- 
  Lars Aronsson (lars at aronsson.se)
  Aronsson Datateknik - http://aronsson.se