Polygons deformed

classic Classic list List threaded Threaded
15 messages Options
Reply | Threaded
Open this post in threaded view
|

Polygons deformed

demon.box
Hi, I see that the polygons are deformed in MapSource compared to the online openstreetmap view.
When I draw a polygon, I make perfect line and angles but when I see it compiled in MapSource it is not the same form.
I don’t understand why.
    Thanks



_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed

osm-8
Am 21.12.2012 23:38, schrieb --enrico:
Hi, I see that the polygons are deformed in MapSource compared to the online openstreetmap view.
When I draw a polygon, I make perfect line and angles but when I see it compiled in MapSource it is not the same form.
I don’t understand why.
    Thanks

This is because of Garmin-map format. Coordinates aren't as precise as OSM format.

Henning

_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed

demon.box
Oh thanks! So there is really nothing to do to solve this problem?
 
Sent: Friday, December 21, 2012 11:39 PM
Subject: Re: [mkgmap-dev] Polygons deformed
 
Am 21.12.2012 23:38, schrieb --enrico:
Hi, I see that the polygons are deformed in MapSource compared to the online openstreetmap view.
When I draw a polygon, I make perfect line and angles but when I see it compiled in MapSource it is not the same form.
I don’t understand why.
    Thanks

This is because of Garmin-map format. Coordinates aren't as precise as OSM format.

Henning


_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed

osm-8
Am 21.12.2012 23:45, schrieb --enrico:
Oh thanks! So there is really nothing to do to solve this problem?
Hi,
I guess mkgmap is rounding the OSM coords. Maybe if them were only inclined or only declined to next Garmin coord there could be more beautiful polygons and ways. But this is only a guess. Maybe someone else can verify this.

Henning

_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed

demon.box
Is a pity that we can not solve this problem. When drawing in OSM I spend a lot of time to trace the building as closely as possible but then I can not see in MapSource.
It seems strange we can not do anything ...
--enrico
 
Sent: Saturday, December 22, 2012 12:30 AM
Subject: Re: [mkgmap-dev] Polygons deformed
 
Am 21.12.2012 23:45, schrieb --enrico:
Oh thanks! So there is really nothing to do to solve this problem?
Hi,
I guess mkgmap is rounding the OSM coords. Maybe if them were only inclined or only declined to next Garmin coord there could be more beautiful polygons and ways. But this is only a guess. Maybe someone else can verify this.

Henning


_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed

Marko Mäkelä
On Sat, Dec 22, 2012 at 09:47:23AM +0100, --enrico wrote:
>Is a pity that we can not solve this problem. When drawing in OSM I
>spend a lot of time to trace the building as closely as possible but
>then I can not see in MapSource.

It is a limitation of the Garmin map format that is supported by mkgmap.  
The map resolution is something like 5 meters, depending on the
latitude.

>It seems strange we can not do anything ...

One thing that mkgmap could do better is to suppress bogus messages
about small roundabouts 'going in wrong direction' when they have been
split to multiple segments due to (bus) route relations.

OsmAnd on an Android device could work for some needs. You have a
variety of screen sizes, but you have to be careful about the GPS
accuracy. Luckily, the Xperia Active hardware has been rather good value
for money for me. Long-distance offline routing is not yet quite there
and there can be some minor glitches.

Best regards,

        Marko
_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed

Chris66
In reply to this post by osm-8
Am 21.12.2012 23:39, schrieb Henning Scholland:

> This is because of Garmin-map format. Coordinates aren't as precise as
> OSM format.

Yes, as far as I know they have a resolution of only 4 meters.

Leading to the funny situation that a roundabout which is mapped
precisely with 100 nodes looks ugly on Garmin compared to a circle
that is mapped with only 8 nodes.

I already experimented with the --reduce-point-density-polygon=NUM
option but without much success.

Chris


_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed

demon.box
I tried with the --reduce-point-density-polygon=NUM but I don't see any
difference....
What a pity this problem!
--enrico

-----Messaggio originale-----
From: Chris66
Sent: Saturday, December 22, 2012 10:03 AM
To: [hidden email]
Subject: Re: [mkgmap-dev] Polygons deformed

Am 21.12.2012 23:39, schrieb Henning Scholland:

> This is because of Garmin-map format. Coordinates aren't as precise as
> OSM format.

Yes, as far as I know they have a resolution of only 4 meters.

Leading to the funny situation that a roundabout which is mapped
precisely with 100 nodes looks ugly on Garmin compared to a circle
that is mapped with only 8 nodes.

I already experimented with the --reduce-point-density-polygon=NUM
option but without much success.

Chris


_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev 

_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed

demon.box
In reply to this post by Chris66
Ideas from someone else?
Thanks!
--enrico

-----Messaggio originale-----
From: Chris66
Sent: Saturday, December 22, 2012 10:03 AM
To: [hidden email]
Subject: Re: [mkgmap-dev] Polygons deformed

Am 21.12.2012 23:39, schrieb Henning Scholland:

> This is because of Garmin-map format. Coordinates aren't as precise as
> OSM format.

Yes, as far as I know they have a resolution of only 4 meters.

Leading to the funny situation that a roundabout which is mapped
precisely with 100 nodes looks ugly on Garmin compared to a circle
that is mapped with only 8 nodes.

I already experimented with the --reduce-point-density-polygon=NUM
option but without much success.

Chris


_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed [Antwort]

Frank Stinner
The problem is the Garmin-Data-Format.

Garmin use only 24 Bit for longitude/latitude. This is for the highest zoom. If you reduce the zoom, then use garmin internaly only 23, 22, ... Bit.
The biggest number with 24 Bit is 16.777.215. Use this for 360°. From this it follows that 360°/16.777.215 is the minimal distance. What means that for example for the equator? 360° stands for around 40.075 km circumference of the earth. The minimal distance is for a latitude of 0° around 2,4m! For higher latitude it is of course higher.

I think, there is no way. 24 Bit for longitude/latitude is to less. But more Bits are more memory consumption and bigger maps.

See also on my web-page "Linien- und Polygonvereinfachung" http://technik.stinnerweb.de/Gps/garmin1.html. It is in german, but the pictures are language-independed ;)

Frank
_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed [Antwort]

Gerd Petermann
Frank Stinner wrote
The problem is the Garmin-Data-Format.

Garmin use only 24 Bit for longitude/latitude. This is for the highest
zoom. If you reduce the zoom, then use garmin internaly only 23, 22, ...
Bit.
The biggest number with 24 Bit is 16.777.215. Use this for 360°. From this
it follows that 360°/16.777.215 is the minimal distance. What means that
for example for the equator? 360° stands for around 40.075 km
circumference of the earth. The minimal distance is for a latitude of 0°
around 2,4m! For higher latitude it is of course higher.

I think, there is no way. 24 Bit for longitude/latitude is to less. But
more Bits are more memory consumption and bigger maps.

See also on my web-page "Linien- und Polygonvereinfachung"
http://technik.stinnerweb.de/Gps/garmin1.html. It is in german, but the
pictures are language-independed ;)
I assume this problem will "disappear" when cheap hardware is able to handle
more details, means, Garmin will sell new hardware with new software and new map
formats or everything will be done just in time using online data.

Gerd
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed [Antwort]

Chris66
In reply to this post by Frank Stinner
Am 24.12.2012 09:26, schrieb Frank Stinner:

> I think, there is no way.

The way would be to improve the reduce-point-density
options.

Chris



_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed [Antwort]

Chris66
Am 24.12.2012 10:53, schrieb Chris66:

> The way would be to improve the reduce-point-density
> options.

Of course this would not help for buildings mapped with 4 points
but could improve roundabouts and streets mapped with a high
density.

Chris


_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed [Antwort]

toc-rox
In reply to this post by Gerd Petermann
GerdP wrote
I assume this problem will "disappear" when cheap hardware is able to handle
more details, means, Garmin will sell new hardware with new software and new map
formats or everything will be done just in time using online data.

Gerd
Offtopic: The competitors are already there ... outdoors mobiles with Android. It would be a life insurance for Garmin to open their format and let the community create full featured, high quality maps. Garmin has a new CEO since a few weaks ... we will see what happens next year.

Klaus
Reply | Threaded
Open this post in threaded view
|

Re: Polygons deformed [Antwort]

Gerd Petermann
In reply to this post by Chris66
Hi,

I think we have solved this problem at least partly during the last weeks.
If you still find deformed polygons or roundabouts that cannot be explained
by the rounding problems, please let me know how to reproduce the problem.
Reg. roundabouts there is one additional source of problems in the remove-short-arcs function
which might "move" the points of the roundabout into the direction of the connected roads.
If this is the case I have an idea how to fix that.

Gerd

Chris66 wrote
Am 24.12.2012 10:53, schrieb Chris66:

> The way would be to improve the reduce-point-density
> options.

Of course this would not help for buildings mapped with 4 points
but could improve roundabouts and streets mapped with a high
density.

Chris


_______________________________________________
mkgmap-dev mailing list
[hidden email]
http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev