Published by breki on 09 Jul 2009 at 09:10 pm
Viewfinder’s DEM Data – Comparison With SRTM
For the past couple of weeks I’ve been busy on extending my digital elevation model (DEM) library with some new features, including the ability to use DEM sources other than the “standard” SRTM3. I’ve managed to produce a working version of the code which supports Viewfinder’s DEM1 data for Alps and I was dying to test it on Julian Alps, especially the area around Triglav, the highest mountain in Slovenia. The SRTM3 data for Julian Alps is very poor, with large chunks of height data missing. Here’s a sample contours map generated in Kosmos using that low-quality SRTM data:
As you can see above, the actual Triglav mountain does not have any contours. Compare that with the contours generated with Viewfinder’s data:
Not only all the contours are there (and seem to correspond with reality), but also the resolution is three times better than that of SRTM3. I’m pretty impressed with Viewfinder’s data, I have to admit.
This is only an intermediate experiment, I have to do some more work on the library before I integrate it into GroundTruth and Kosmos. But the results so far are very promising.
BTW: I contacted Jonathan de Ferranti, the author of Viewfinder’s site, and asked him about terms of use of these DEM tiles. Basically he gave a wholehearted approval of using it for open-source projects as long as the original authorship is respected. He does not claim copyright on these tiles since they originated from various sources, be he still likes to be informed about any large-scale use of them.
sim on 10 Jul 2009 at 4:20 #
just a note, some of the Viewfinder data stems from Soviet military topos, these a copyrighted by Roscartographia, still. No reproduction, no selling etc. We’re not using these data for open projects, it is avoided for OSM etc…
I know you’re guys in the different country, so you don’t probably care, but this kind of permission sounds to me a little like if someone person will allow you to use his neighbour’s car.
breki on 10 Jul 2009 at 5:58 #
@sim,
I do care about copyright, regardless of the country of origin. If there are any copyright issues, the data will not be used (and I’ll probably disable support for such a source in my DEM library).
As for the “neighbour’s car”: Viewfinder’s author did explicitly state he doesn’t hold the copyright to the source data, but he didn’t receive any objections in the 4 years these DEMs have been available. Have you contacted him about the Roscartographia?
sim on 10 Jul 2009 at 6:46 #
well, it is not like I want to enforce their copyright. In fact I think this particular one is stupid. I might drop author a line, so that he would know.
Navigator on 03 Aug 2009 at 0:32 #
Hi,
have also experimented with those Viewfinder datasets a few months ago. Better resolution for sure, but often unrealistic contours. Wouldn’t it make sense to implement the new ASTER-Datasets and kick the SRTMs away? They have 1 arc-secong giving a 30 meters resolution. But i think they can’t be directly downloaded online to programs like Kosmos. A registration is needed to get the datasets for free. Would be great if they could be used with Kosmos.
More informations and comparisons of the resolution: http://forums.atomic-systems.com/viewtopic.php?f=304&t=96025&start=15 http://forums1.avsim.net/lofiversion/index.php/t253387.html
Informations, Registration and Data-download: http://asterweb.jpl.nasa.gov/OnDemand.asp https://wist.echo.nasa.gov/~wist/api/imswelcome/ https://wist.echo.nasa.gov/wist-bin/api/ims.cgi?mode=MAINSRCH&JS=1
breki on 03 Aug 2009 at 7:54 #
@Navigator,
First of all, thanks for the links. I’ve had a brief looked into ASTER when they released it a few months ago, but the fact you had to be registered in order to download the data put me off it, at least for now. SRTM is not very good, but at least it’s easily available for any user to access (and any software to download it transparently), so ASTER cannot be a replacement for it.
Also, from what I know, the terms of use for ASTER are a bit problematic from the OSM perspective, so it’s not as clean as “public domain”.
I’ll try to add ASTER in the future, but right now it’s not on the priority list.
Andy Allan on 18 Jan 2010 at 15:34 #
If you’re interested in ASTER and OpenStreetMap, you might want to read about my experiences with it.
http://www.gravitystorm.co.uk/shine/archives/2010/01/13/aster-not-worth-it-yet/
breki on 18 Jan 2010 at 18:22 #
@Andy,
Thanks for the link. From your post, ASTER actually looks worse than I initially thought. I guess SRTM is still the only option.