Documentation for the new Douglas-Peucker options in low-res-opt branch
Hi all, I've now added documentation for these new options, see: https://www.mkgmap.org.uk/websvn/diff.php?repname=mkgmap&path=%2Fbranches%2F... Is it clear enough? I think the recommend value --reduce-point-density-polygon=8 is far too high at low resolutions. Should this be changed? Together with the new --improve-overview option this branch version can produce much better results for the lower resolutions. Gerd
Hi Gerd, I think the original documentation for --reduce-point-density and --reduce-point-density-polygon could do with some improvement. It also seems bizarre to have a recommended value that is not the default. Is 2.6 the default if --reduce-point-density is specified without a value, or is it also the default if the option is not specified? Are the units metres? Is the distance the same no matter what resolution is used, or does the distance increase at lower resolution? If the former, wouldn't it be better to increase by a factor of 1.414 at successive resolutions? Would this be sufficient to not need to be able to specify individual values for resolutions? I'm not keen on having two very differently named options that basically achieve the same aim and suggest that it would be better to simply extend the existing --reduce-point-density options with --reduce-point-density=value|resolution:value[,...] or even better --reduce-point-density=value[,...] where the first value applies to the first used resolution and so on, with the last value being scaled for any further resolutions that have not had a value specified. Is there a reason why polygons need different values than lines? Shouldn't reduce-point-density-polygon default to the reduce-point-density value? I note that although the documentation belongs to the low-res-branch, it is showing up on the mkgmap command line web page. Regards, Mike -----Original Message----- From: Gerd Petermann [mailto:GPetermann_muenchen@hotmail.com] Sent: 14 June 2021 07:43 To: mkgmap-dev@lists.mkgmap.org.uk Subject: [mkgmap-dev] Documentation for the new Douglas-Peucker options in low-res-opt branch Hi all, I've now added documentation for these new options, see: https://www.mkgmap.org.uk/websvn/diff.php?repname=mkgmap&path=%2Fbranches%2F low-res-opt%2Fresources%2Fhelp%2Fen%2Foptions&rev=4775&peg=4775 Is it clear enough? I think the recommend value --reduce-point-density-polygon=8 is far too high at low resolutions. Should this be changed? Together with the new --improve-overview option this branch version can produce much better results for the lower resolutions. Gerd
I think polygons can be higher value except at level=0. In general I think it is better to have a bit higher values for higher levels. However then again lower for the overview map. However if the improve-overview filter chain would apply in general - then values need to be much lower. There is very aggressive filtering in comparison. -x-simplify-filter-line-errors=23:2.6,22:4.2,21:5.4,20:6,19:7,18:7.5,17:5.2,13:5.2 --x-simplify-filter-polygon-errors=23:3.6,22:7,21:6,20:9,17:2.6 Why do we need higher values for resolutions far zoomed out? Most garmin GPS devices are pretty slow if zoomed out far - so heavily filter those polygons/lines to get decent speed. PC/desktop are fast enough for even complicated overview map - so that one should be optimized for best visual quality. Some filtering is needed usually however else Asia continent map or Europe continent map would break the img size limit. This is what I am using now - and it works pretty well. Also the polygon size limits I feel the following is way better than default. Tiny dots look more like errors than help much: --polygon-size-limits="24:12, 23:14, 22:14, 21:20, 20:20, 19:20, 18:20, 17:20, 16:20, 15:20, 14:20, 13:25" On Mon, 14 Jun 2021 at 23:32, Mike Baggaley <mike@tvage.co.uk> wrote:
Hi Gerd,
I think the original documentation for --reduce-point-density and --reduce-point-density-polygon could do with some improvement. It also seems bizarre to have a recommended value that is not the default. Is 2.6 the default if --reduce-point-density is specified without a value, or is it also the default if the option is not specified? Are the units metres? Is the distance the same no matter what resolution is used, or does the distance increase at lower resolution? If the former, wouldn't it be better to increase by a factor of 1.414 at successive resolutions? Would this be sufficient to not need to be able to specify individual values for resolutions?
I'm not keen on having two very differently named options that basically achieve the same aim and suggest that it would be better to simply extend the existing --reduce-point-density options with --reduce-point-density=value|resolution:value[,...] or even better --reduce-point-density=value[,...] where the first value applies to the first used resolution and so on, with the last value being scaled for any further resolutions that have not had a value specified.
Is there a reason why polygons need different values than lines? Shouldn't reduce-point-density-polygon default to the reduce-point-density value?
I note that although the documentation belongs to the low-res-branch, it is showing up on the mkgmap command line web page.
Regards, Mike
-----Original Message----- From: Gerd Petermann [mailto:GPetermann_muenchen@hotmail.com] Sent: 14 June 2021 07:43 To: mkgmap-dev@lists.mkgmap.org.uk Subject: [mkgmap-dev] Documentation for the new Douglas-Peucker options in low-res-opt branch
Hi all,
I've now added documentation for these new options, see:
https://www.mkgmap.org.uk/websvn/diff.php?repname=mkgmap&path=%2Fbranches%2F low-res-opt%2Fresources%2Fhelp%2Fen%2Foptions&rev=4775&peg=4775 <https://www.mkgmap.org.uk/websvn/diff.php?repname=mkgmap&path=%2Fbranches%2Flow-res-opt%2Fresources%2Fhelp%2Fen%2Foptions&rev=4775&peg=4775>
Is it clear enough? I think the recommend value --reduce-point-density-polygon=8 is far too high at low resolutions. Should this be changed?
Together with the new --improve-overview option this branch version can produce much better results for the lower resolutions.
Gerd
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
-- Felix Hartman - Openmtbmap.org & VeloMap.org
Hi Mike, some answers: - the unit is metre, it is scaled with this formular: maxErrorDistance = filterDistance * (1<< (24 - resolution)), so factor 4 for resolution 22, 8 for 21, 16 for 20 and so on. - The default is 2.6 if the option is not given or given without a value. The --no prefix doesn't work, but the Douglas-Peucker filter can be switched off with a value of 0. - I used different option names because I think --reduce-point-density is a bit misleading. I didn't think about changing the syntax of the existing option, but it would be an alternative. - Why different option for polygons? See http://gis.19327.n8.nabble.com/Include-the-following-patches-into-trunk-Patc... I really have no clue how one (Felix) determines his favorite values and it's not interesting for me as a software developer. @Steve Reg. online doc: I don't know if this is intended. I'd prefer to have separate docs for each branch, or that branch builds don't change the web page, whatever is easier. Gerd ________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Mike Baggaley <mike@tvage.co.uk> Gesendet: Montag, 14. Juni 2021 22:32 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] Documentation for the new Douglas-Peucker options in low-res-opt branch Hi Gerd, I think the original documentation for --reduce-point-density and --reduce-point-density-polygon could do with some improvement. It also seems bizarre to have a recommended value that is not the default. Is 2.6 the default if --reduce-point-density is specified without a value, or is it also the default if the option is not specified? Are the units metres? Is the distance the same no matter what resolution is used, or does the distance increase at lower resolution? If the former, wouldn't it be better to increase by a factor of 1.414 at successive resolutions? Would this be sufficient to not need to be able to specify individual values for resolutions? I'm not keen on having two very differently named options that basically achieve the same aim and suggest that it would be better to simply extend the existing --reduce-point-density options with --reduce-point-density=value|resolution:value[,...] or even better --reduce-point-density=value[,...] where the first value applies to the first used resolution and so on, with the last value being scaled for any further resolutions that have not had a value specified. Is there a reason why polygons need different values than lines? Shouldn't reduce-point-density-polygon default to the reduce-point-density value? I note that although the documentation belongs to the low-res-branch, it is showing up on the mkgmap command line web page. Regards, Mike -----Original Message----- From: Gerd Petermann [mailto:GPetermann_muenchen@hotmail.com] Sent: 14 June 2021 07:43 To: mkgmap-dev@lists.mkgmap.org.uk Subject: [mkgmap-dev] Documentation for the new Douglas-Peucker options in low-res-opt branch Hi all, I've now added documentation for these new options, see: https://www.mkgmap.org.uk/websvn/diff.php?repname=mkgmap&path=%2Fbranches%2F low-res-opt%2Fresources%2Fhelp%2Fen%2Foptions&rev=4775&peg=4775 Is it clear enough? I think the recommend value --reduce-point-density-polygon=8 is far too high at low resolutions. Should this be changed? Together with the new --improve-overview option this branch version can produce much better results for the lower resolutions. Gerd _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
I just compile a map of usually Austria or Switzerland, use different values and first look at it in Basecamp. If it looks good enough I then transfer it to both Oregon 600 and etrex 30x and look at the map on the device, cycling through all zoom levels and panning the map around. On any not really old PC the best value would be very small of course, as there was is no speed problem. So the biggest value that looks good is preferably. However on the GPS device such a low value for resolution 21-18 is pretty slow with most map styles. So a compromise has to be found between higher values or showing less information. Except for automobile use, resolution 21-18 is only used for getting an overview, meaning panning the map around. If it is too slow it becomes useless. That is why I use comparatively high values for those resolutions. A map to be used in a car needs lower values at 20 and 21 at least. Polygons like forest, lakes and so on when zoomed out far don't need to be very exact, so higher values compared to lines make sense. Just not at level 0 where higher values lead to houses not being rectangular anymore On Tue, 15 Jun 2021, 07:36 Gerd Petermann <gpetermann_muenchen@hotmail.com> wrote:
Hi Mike,
some answers: - the unit is metre, it is scaled with this formular: maxErrorDistance = filterDistance * (1<< (24 - resolution)), so factor 4 for resolution 22, 8 for 21, 16 for 20 and so on. - The default is 2.6 if the option is not given or given without a value. The --no prefix doesn't work, but the Douglas-Peucker filter can be switched off with a value of 0. - I used different option names because I think --reduce-point-density is a bit misleading. I didn't think about changing the syntax of the existing option, but it would be an alternative. - Why different option for polygons? See http://gis.19327.n8.nabble.com/Include-the-following-patches-into-trunk-Patc...
I really have no clue how one (Felix) determines his favorite values and it's not interesting for me as a software developer.
@Steve Reg. online doc: I don't know if this is intended. I'd prefer to have separate docs for each branch, or that branch builds don't change the web page, whatever is easier.
Gerd
________________________________________ Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Mike Baggaley <mike@tvage.co.uk> Gesendet: Montag, 14. Juni 2021 22:32 An: mkgmap-dev@lists.mkgmap.org.uk Betreff: Re: [mkgmap-dev] Documentation for the new Douglas-Peucker options in low-res-opt branch
Hi Gerd,
I think the original documentation for --reduce-point-density and --reduce-point-density-polygon could do with some improvement. It also seems bizarre to have a recommended value that is not the default. Is 2.6 the default if --reduce-point-density is specified without a value, or is it also the default if the option is not specified? Are the units metres? Is the distance the same no matter what resolution is used, or does the distance increase at lower resolution? If the former, wouldn't it be better to increase by a factor of 1.414 at successive resolutions? Would this be sufficient to not need to be able to specify individual values for resolutions?
I'm not keen on having two very differently named options that basically achieve the same aim and suggest that it would be better to simply extend the existing --reduce-point-density options with --reduce-point-density=value|resolution:value[,...] or even better --reduce-point-density=value[,...] where the first value applies to the first used resolution and so on, with the last value being scaled for any further resolutions that have not had a value specified.
Is there a reason why polygons need different values than lines? Shouldn't reduce-point-density-polygon default to the reduce-point-density value?
I note that although the documentation belongs to the low-res-branch, it is showing up on the mkgmap command line web page.
Regards, Mike
-----Original Message----- From: Gerd Petermann [mailto:GPetermann_muenchen@hotmail.com] Sent: 14 June 2021 07:43 To: mkgmap-dev@lists.mkgmap.org.uk Subject: [mkgmap-dev] Documentation for the new Douglas-Peucker options in low-res-opt branch
Hi all,
I've now added documentation for these new options, see:
https://www.mkgmap.org.uk/websvn/diff.php?repname=mkgmap&path=%2Fbranches%2F low-res-opt%2Fresources%2Fhelp%2Fen%2Foptions&rev=4775&peg=4775 <https://www.mkgmap.org.uk/websvn/diff.php?repname=mkgmap&path=%2Fbranches%2Flow-res-opt%2Fresources%2Fhelp%2Fen%2Foptions&rev=4775&peg=4775>
Is it clear enough? I think the recommend value --reduce-point-density-polygon=8 is far too high at low resolutions. Should this be changed?
Together with the new --improve-overview option this branch version can produce much better results for the lower resolutions.
Gerd
_______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev _______________________________________________ mkgmap-dev mailing list mkgmap-dev@lists.mkgmap.org.uk https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
participants (4)
-
Felix Hartmann -
Gerd Petermann -
Gerd Petermann
-
Mike Baggaley