[mkgmap-dev] DEM: performance with bicubic interpolation

Tue Jan 23 14:23:34 GMT 2018

Maybe I'm thinking too scientific about it, but what benefit should be
expected when using lower --dem-dist values then the source can provide?
The interpolation is finally done by the Garmin software while
displaying the elevation data. At least I'm getting more elevation data
points in Basecamp then in the map. I'm using --dem-dist=9942 for 3"
srtm and at least elevation plot in routing shows distances between
points of less than 90m. Also while moving the mouse you can see a
continues reduction of altitude. There are no jumps in mountainous area.

So I don't see a reason to read 'rough' source data, making them
theoretical more accurate (which isn't correct anyway, it's more
guessing than knowing) and finally the interpolated values are
interpolated again. Maybe it looks smoother, but it shouldn't be more
accurate. So I doubt it's worth.

Henning

On 23.01.2018 22:05, Gerd Petermann wrote:
> Hi Henning,
> yes, we have to scale hgt resolution to garmin resolution, this is not about voids.
>
> We may use the same algo to fill voids. I think this link is quite interesting:
> https://web.archive.org/web/20051024202307/http://www.geovista.psu.edu/sites/geocomp99/Gc99/082/gc_082.htm
>
> Gerd
>
>
>
> ________________________________________
> Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von Henning Scholland <osm at hscholland.de>
> Gesendet: Dienstag, 23. Januar 2018 14:09:17
> An: mkgmap-dev at lists.mkgmap.org.uk
> Betreff: Re: [mkgmap-dev] DEM: performance with bicubic interpolation
>
> For srtm I found elevation accuracy of 6m in wikipedia, so my 10m was not that unrealistic.
>
> Btw. We are talking about interpolating from hgt-data position to Garmin-data position, aren't we?
>
> For filling voids, it would be a good idea to use some spline interpolation as written here:
> https://en.wikipedia.org/wiki/Shuttle_Radar_Topography_Mission#Void-filled_SRTM_datasets
>
> Henning
>
>
> On 23.01.2018 20:59, Henning Scholland wrote:
>
> Hi Andrzej,
>
> I also suggest to make interpolation optional.
>
> So far I don't understand your argument. I agree, compilation time is
> not the only criteria. The question is, what is the benefit for the user.
> For example: If accuracy of srtm is +- 10m and the difference between
> with/without interpolation is +-1m, then it's definitely not worth
> spending any effort on interpolation. Only in cases, where interpolation
> accuracy is on same level than srtm accuracy, it starting to be worth
> spending time for it.
>
> Do you have any values for differences with/without interpolation with
> same input data?
>
> Henning
>
> On 23.01.2018 20:22, Andrzej Popowski wrote:
>
>
> Hi Gerd,
>
> there are different kind of bicubic interpolations. I'm not good at
> this math, I think the previous version was actually bicubic spline
> interpolation. See other possibilities here:
> http://mrl.nyu.edu/~perlin/cubic/Cubic_java.html
>
> I don't know, which type of spline or cubic interpolation is best for
> DEM. Maybe differences aren't big, but I think it is better to have
> good interpolation than fast one. You compile a map once but then many
> people can use it.
>
> It shouldn't be difficult to include an option like:
> --dem-interp=..
>
>
>
>
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-dev at lists.mkgmap.org.uk<mailto: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
>