| intersection of two tracks(lat,lon) [message #75069] |
Wed, 16 February 2011 00:59  |
Kai Muehlbauer
Messages: 13
Registered: February 2011
|
Junior Member |
|
|
Hi,
I've consulted already the usual IDL resources. Maybe I overlooked
something.
I'm looking for a program which calculates the intersection points for
lines or polygon edges similar to the matlab function "POLYXPOLY".
Aim is to find the intersection (Lat,Lon) of two Tracks (Latitude,
Longitude). Also the index of the intersection point is needed.
Any hint which can direct me to an solution is appreciated.
Cheers
Kai
|
|
|
|
|
|
|
|
|
|
|
|
| Re: intersection of two tracks(lat,lon) [message #75213 is a reply to message #75069] |
Mon, 21 February 2011 22:51  |
Kai Muehlbauer
Messages: 13
Registered: February 2011
|
Junior Member |
|
|
Hi all,
Am 21.02.2011 15:05, schrieb Guillermo:
> So the fact that you are still considering the other solution makes
> me curious, what in earth are you tracking??
Uhm, not real tracking, it is actually the intersection of two radar
beams (center of beam).
Cheers
Kai
|
|
|
|
| Re: intersection of two tracks(lat,lon) [message #75232 is a reply to message #75069] |
Mon, 21 February 2011 06:05 |
guillermo.castilla.ca
Messages: 27
Registered: September 2008
|
Junior Member |
|
|
> Am 17.02.2011 09:22, schrieb Kai Muehlbauer:
>
> Actually I do (as Guillermo suggested) check the heights of the two
> tracks at intersection point. This is sufficiently fast and suits fine.
Then why make it complicated by going for the brute force solution??
The latter would only make sense if your tracks consisted of GPS
waypoints measured say every millisecond. Otherwise the fact that two
waypoints from different tracks are close (i.e., a local minimum in
the distance array) is no guarantee that the tracks actually intersect
(they could run in parallel and be a narrowing in that section).
Conversely, if the tracks are measured at different time lags, or if
the objects have very different speed, an intersection does not need
to translate into a local minimum in the distance array (the waypoints
could be far away from the intersection). So the fact that you are
still considering the other solution makes me curious, what in earth
are you tracking??
Cheers
Guillermo
|
|
|
|
| Re: intersection of two tracks(lat,lon) [message #75238 is a reply to message #75136] |
Sun, 20 February 2011 23:59 |
Kai Muehlbauer
Messages: 13
Registered: February 2011
|
Junior Member |
|
|
Hi all,
Am 17.02.2011 09:22, schrieb Kai Muehlbauer:
> Unfortunately, this is only for X,Y-space. What can I do, if the 3rd
> dimension (height) should also be processed?
my thanks to Guillermo and also Ken for sharing your ideas.
Actually I do (as Guillermo suggested) check the heights of the two
tracks at intersection point. This is sufficiently fast and suits fine.
Also Kens suggestion of a brute force solution finding the distances for
every pair of points is interesting. As I'am not an idl-expert I
appreciate some coding-help at this point. I found that I could use
DISTANCE_MEASURE, but it has only one array as input argument. So I have
to concatenate my two (x,y,z)-tracks and after processing extract the
interesting indices from the output array/matrix. Then I have to sort,
to find the shortest distance (or do some more sophisticated) and
possible local minima.
Cheers,
Kai
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
