logo separator

[mkgmap-dev] Tiles pruned in DEM map

From Gerd Petermann gpetermann_muenchen at hotmail.com on Sun Dec 27 09:36:03 GMT 2020

Hi Carlos,

I agree that --order-by-decreasing-area is to blame. This option seems to disturb the calculation of the overview map, and maybe the extreme size of the tile plays another role here.

I hope Ticker can help here, I think the option supresses the merging of shapes, and this is probably not a good idea for the 0x4a shapes.

As a work-around for you:
Add option --no-order-by-decreasing-area at the end(!) of the parameter list so that the overview map is created without this option. This seems to fix the problem on my machine.

Gerd

________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von Gerd Petermann <gpetermann_muenchen at hotmail.com>
Gesendet: Sonntag, 27. Dezember 2020 09:36
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] Tiles pruned in DEM map

Hi Carlos,

reg. the routing data: You can run display tool to check yourself, but the error is either in mkgmap or in display tool. My command looks like this :
java -ea -cp d:\display\dist\display.jar;d:\mkgmap\dist\mkgmap.jar test.check.NodCheck  -vv --tab-zero=0 51145001.img > nodcheck.51145001.img 2> crash_nod
It would be easier if you would post a link to your style.

I am now able to reproduce the display error using this command
java -Xmx6G -ea -jar d:\mkgmap\dist\mkgmap.jar --dem=f:\srtm3 --dem-poly=e:\carlos\australia.poly  --overview-dem-dist=276160 --gmapi --show-profiles=1 --bounds=f:\osm\bounds.zip --precomp-sea=f:\osm\sea.zip --route --reduce-point-density=4 --polygon-size-limits="24:12, 18:10, 16:0" --order-by-decreasing-area --location-autofill=is_in,nearest --drive-on=detect,left --road-name-config=d:\mkgmap\resources\roadNameConfig.txt  e:\carlos\51145001.o5m

and the overview map has the same amount of 0x4a and 0x4b polygons as yours from 11-error.zip.

Gerd

________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von Carlos Dávila <carlos at alternativaslibres.org>
Gesendet: Samstag, 26. Dezember 2020 19:12
An: mkgmap-dev at lists.mkgmap.org.uk
Betreff: Re: [mkgmap-dev] Tiles pruned in DEM map

I'm sorry for the late response and for breaking the thread, but I
didn't receive Gerd's last message, so I've had to copy it from mailing
list archive. Reply inline.

OK, I think I understand what problem you see. I used JaVaWa
MapConverter to install the map in 11-error.zip and 12-OK.zip played
with it. With 11-error.zip only the data from the overview map is
displayed when I zoom to e.g. S21.71088 E114.92708, while 12-OK.zip
shows more details (this is with MapSource, in Basecamp I see details in
both maps).

Yes, everything to the North of approx. S32.39125 is missing.

I tried to analyse your files and I see three suspicious things so far:

1) The routing data seems to be wrong, NodCheck reports "Could not find
node for road 38a61 nod2=124c8 " for both tiles. I don't see such an
error with the default style.

How can I check that? If both tiles are affected, probably this is not
related with current issue, but other maps produced with the same style
will also be wrong regarding routing.

2) The bad overview map contains a lot more 0x4a polygons. This is
probably caused by the --order-by-decreasing-area, and I am not sure why
this happens.

Do you think the problem is in the overview map or may be in tile map?
If I open tile in MapEdit++ the number of non polygon elements is almost
the same in wrong and correct maps. For example, number of roads is
exactly the same. It seems as if something is masking part of the tile
in MapSource (also in BaseCamp in my case); elements are there, but you
can't see them.

3) You use a special version of mkgmap, so please try also with the
latest version.

With latest mkgmap and default style problem persists.

My first guess was that the bad NOD file may cause this but the problem
doesn't disappear when I remove the file 51145001.NOD, so this is
probably not to blame. Same for the 51145001.DEM file.

I tried to reproduce the possible problem with the 0x4a tiles using the
default style with this command java -Xmx4G -jar map.jar --route --gmapi
--bounds=f:\osm\bounds.zip --precomp-sea=f:\osm\sea.zip
--reduce-point-density=4 --polygon-size-limits="24:12, 18:10, 16:0"
--order-by-decreasing-area --location-autofill=is_in,nearest
f:\dwnload\temp\51145001.o5m but my overview map contains the same
number of 0x4a tiles as your good map.

I cannot reproduce your DEM data because I don't know the polygon file
(polygons/$pais.poly). Maybe I can reproduce the 0x4a problem when you
create the test files again without --remove-ovm-work-files=true.

You can download polygon file from here
<https://alternativaslibres.org/tmp/australia.poly> and ovm from here
<https://alternativaslibres.org/tmp/ovm_51145001.img>.

El 22/12/20 a las 18:05, Carlos Dávila escribió:
> I'm sorry, probably I didn't explain well enough.
>
> Overview is always correct, the problem affects only tiles. As you saw
> in the screenshots of my first mail, they are different in size, but
> they are generated from the same input files, so they should have the
> same size. If you zoom in to an area that should be included in a
> tile, only overview map is shown, no detailed map. Even more, when you
> zoom in, at the given point where detailed map appears, tile boundary
> jumps to the correct size, but nothing but overview is displayed
> outside the "pruned" tile.
>
> You can download correct and wrong files from the links below and play
> with them to get a better idea of the problem. They correspond to
> first tile of Australia as shown in my screenshots.
>
> https://alternativaslibres.org/tmp/11-error.zip
>
> https://alternativaslibres.org/tmp/12-OK.zip
>
> Error was generated with java -Xmx27G -ea
> -Dlog.config=logging.properties -jar mkgmap.jar
> --dem=hgt/LIDAR-Sonny,hgt/ALOS,hgt/VIEW1,hgt/SRTM1,hgt/EU-DEM,hgt/VIEW3,hgt
> --dem-poly=polygons/$pais.poly --overview-dem-dist=276160 -c
> opciones_comunes $CODEPAGE --gmapi --show-profiles=1
> --bounds=bounds.zip --precomp-sea=sea.zip --route --family-name="OSM
> $MAPA DEM" --family-id=5${FID} --overview-mapname=${ABR}5${FID}
> --overview-mapnumber=${GRUPO}5${FID}000 --reduce-point-density=4
> --polygon-size-limits="24:12, 18:10, 16:0" --order-by-decreasing-area
> --location-autofill=is_in,nearest --drive-on=detect,$DRIVEON $LANGUAGE
> --road-name-config=${CONFIG} --style=mio --remove-ovm-work-files=true
> 51145001.o5m tmp/${ABR}5${FID}.txt
>
> OK was generated with java -Xmx27G -ea -Dlog.config=logging.properties
> -jar mkgmap.jar
> --dem=hgt/LIDAR-Sonny,hgt/ALOS,hgt/VIEW1,hgt/SRTM1,hgt/EU-DEM,hgt/VIEW3,hgt
> --dem-poly=polygons/$pais.poly --overview-dem-dist=276160 -c
> opciones_comunes $CODEPAGE --gmapi --show-profiles=1
> --bounds=bounds.zip --precomp-sea=sea.zip --route --family-name="OSM
> $MAPA DEM" --family-id=5${FID} --overview-mapname=${ABR}5${FID}
> --overview-mapnumber=${GRUPO}5${FID}000 --reduce-point-density=4
> --polygon-size-limits="24:12, 18:10, 16:0"
> --location-autofill=is_in,nearest --drive-on=detect,$DRIVEON $LANGUAGE
> --road-name-config=${CONFIG} --style=mio --remove-ovm-work-files=true
> 51145001.o5m tmp/${ABR}5${FID}.txt
>
> Although removing --overview-dem-dist solved the issue in my first
> test (see command below, correct output), after a lot of tests
> removing options one by one from the original command, finally it
> seems the problem is caused by option --order-by-decreasing-area (or
> the combination of both).
>
> java -Xmx27G -ea -Dlog.config=logging.properties -jar mkgmap.jar
> --dem=hgt/LIDAR-Sonny,hgt/ALOS,hgt/VIEW1,hgt/SRTM1,hgt/EU-DEM,hgt/VIEW3,hgt
> --dem-poly=polygons/$pais.poly -c opciones_comunes $CODEPAGE --gmapi
> --show-profiles=1 --bounds=bounds.zip --precomp-sea=sea.zip --route
> --country-name=$PAIS --country-abbr=${ABR} --area-name="$MAPA"
> --family-name="OSM $MAPA DEM" --family-id=5${FID}
> --product-version=$VERSION --series-name="OSM-$MAPA-DEM"
> --overview-mapname=${ABR}5${FID}
> --overview-mapnumber=${GRUPO}5${FID}000 --name-tag-list=$NAMETAG
> --process-destination --process-exits --housenumbers
> --reduce-point-density=4 --polygon-size-limits="24:12, 18:10, 16:0"
> --order-by-decreasing-area --add-pois-to-areas --link-pois-to-ways
> --location-autofill=is_in,nearest --drive-on=detect,$DRIVEON
> --check-roundabouts --check-roundabout-flares
> --license-file=license_OSM_${HGT} --copyright-file=copyright $LANGUAGE
> --road-name-config=${CONFIG} --style=mio --remove-ovm-work-files=true
> 51145001.o5m tmp/${ABR}5${FID}.txt
>
> El 22/12/20 a las 10:15, Gerd Petermann escribió:
>> Hi Carlos,
>>
>> It seems I still don't understand what the problem is or how to
>> reproduce it.
>> I tried this command and the overview map looks OK:
>> java -Xmx4G -jar d:\mkgmap\dist\mkgmap.jar --dem=f:\srtm3
>> --overview-dem-dist=128000 --show-profiles=1 --gmapi
>> f:\dwnload\temp\51145001.o5m
>>
>> If it is related to the DEM data I probably cannot help much. You may
>> try a slightly different value for the --overview-dem-dist option or
>> a modified polygon
>> given with the --dem-poly option.
>>
>> Gerd
>>
>> ________________________________________
>> Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag
>> von Carlos Dávila <carlos at alternativaslibres.org>
>> Gesendet: Montag, 21. Dezember 2020 22:09
>> An: mkgmap-dev at lists.mkgmap.org.uk
>> Betreff: Re: [mkgmap-dev] Tiles pruned in DEM map
>>
>> The problem seems to be caused by overview DEM. If I remove
>> --overview-dem-dist option tile is complete in MapSource. The issue can
>> be reproduced with this tile
>> <https://alternativaslibres.org/tmp/51145001.o5m> and HGT from
>> vierfinderpanoramas3
>>
>> El 18/12/20 a las 11:48, Gerd Petermann escribió:
>>> Hi Carlos,
>>>
>>> not sure if I understand what the problem is. The two screenshots
>>> show completely different tile boundaries, so they were not created
>>> from the same splitter output.
>>>
>>> Gerd
>>>
>>> ________________________________________
>>> Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag
>>> von Carlos Dávila <cardavilam at gmail.com>
>>> Gesendet: Donnerstag, 17. Dezember 2020 23:44
>>> An: Development list for mkgmap
>>> Betreff: [mkgmap-dev] Tiles pruned in DEM map
>>>
>>> I build several types of map (OSM, OSM+contour lines, maps for trucks
>>> and OSM+DEM) for the same area, using same input files. I split
>>> country.o5m with splitter and then use the same template.args output by
>>> splitter for all maps, just changing mapname for the different types of
>>> map. Given that, all resulting mapsets should have the same tiles, but
>>> tiles in DEM map are smaller than in the other types. They share
>>> part of
>>> the boundaries (usually two of them) with other types, but other
>>> boundaries are moved in, reducing displayed tile size. See attached
>>> screenshots. However, file size is the same (discounting *.DEM) for
>>> each
>>> tile in *.gmap subfolders.
>>>
>>> Command is quite similar for OSM and OSM+DEM maps:
>>>
>>> java -Xmx27G -ea -Dlog.config=logging.properties -jar mkgmap.jar -c
>>> opciones_comunes $CODEPAGE --gmapi --bounds=bounds.zip
>>> --precomp-sea=sea.zip --route --country-name=$PAIS
>>> --country-abbr=${ABR}
>>> --area-name=$MAPA --family-name="OpenStreetMap $MAPA"
>>> --family-id=1${FID} --product-version=$VERSION
>>> --series-name="OSM-$MAPA"
>>> --overview-mapname=${ABR}1${FID}
>>> --overview-mapnumber=${GRUPO}1${FID}000
>>> --name-tag-list=$NAMETAG --process-destination --process-exits
>>> --housenumbers --reduce-point-density=4 --polygon-size-limits="24:12,
>>> 18:10, 16:0" --order-by-decreasing-area --add-pois-to-areas
>>> --report-similar-arcs --report-dead-ends=2 --link-pois-to-ways
>>> --location-autofill=is_in,nearest --drive-on=detect,$DRIVEON
>>> --check-roundabouts --check-roundabout-flares
>>> --license-file=license_OSM
>>> --copyright-file=copyright $LANGUAGE --road-name-config=${CONFIG}
>>> --style=mio --check-styles -c $pais.args tmp/${ABR}1${FID}.txt
>>>
>>> java --Xmx27G -ea -Dlog.config=logging.properties -jar mkgmap.jar
>>> --dem=hgt/LIDAR,hgt/ALOS,hgt/VIEW1,hgt/SRTM1,hgt/EU-DEM,hgt/VIEW3,hgt
>>> --dem-poly=polygons/$pais.poly --dem-dists=3312,6624,9936,13248,26512
>>> --overview-dem-dist=128000 -c opciones_comunes $CODEPAGE --gmapi
>>> --show-profiles=1 --bounds=bounds.zip --precomp-sea=sea.zip --route
>>> --country-name=$PAIS --country-abbr=${ABR} --area-name="$MAPA"
>>> --family-name="OSM $MAPA DEM" --family-id=5${FID}
>>> --product-version=$VERSION --series-name="OSM-$MAPA-DEM"
>>> --overview-mapname=${ABR}5${FID}
>>> --overview-mapnumber=${GRUPO}5${FID}000
>>> --name-tag-list=$NAMETAG --process-destination --process-exits
>>> --housenumbers --reduce-point-density=4 --polygon-size-limits="24:12,
>>> 18:10, 16:0" --order-by-decreasing-area --add-pois-to-areas
>>> --link-pois-to-ways --location-autofill=is_in,nearest
>>> --drive-on=detect,$DRIVEON --check-roundabouts
>>> --check-roundabout-flares
>>> --license-file=license_OSM_${HGT} --copyright-file=copyright $LANGUAGE
>>> --road-name-config=${CONFIG} --style=mio
>>> --remove-ovm-work-files=true -c
>>> ${pais}-DEM.args tmp/${ABR}5${FID}.txt
>>>
>>> Both OSM and OSM+DEM maps can be downloaded from
>>> https://alternativaslibres.org/en/downloads.php#Oceania
>>>
>>> Any idea why this happens?
>>>
>>> _______________________________________________
>>> mkgmap-dev mailing list
>>> mkgmap-dev at lists.mkgmap.org.uk
>>> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
>> _______________________________________________
>> mkgmap-dev mailing list
>> mkgmap-dev at lists.mkgmap.org.uk
>> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
>> _______________________________________________
>> mkgmap-dev mailing list
>> mkgmap-dev at lists.mkgmap.org.uk
>> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-dev at lists.mkgmap.org.uk
> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev


More information about the mkgmap-dev mailing list