|
»
Black holes in hillshade image
|
View:
New views
5 Messages
—
Rating Filter:
Alert me
|
 |
Black holes in hillshade image
by Adam Ratcliffe-2
::
Rate this Message:
I've created a hillshade image using the NASA SRTM3 data following the notes at: http://wiki.openstreetmap.org/wiki/HikingBikingMaps. Looks great except for the occasional 'black hole' in the image, I expect this is due to data voids not being rendered correctly.
What's got me puzzled about this is that the wiki page suggests using a value of 32767 for the -srcnodata and -dstnodata options to gdalwarp, howver in the NASA docs (http://dds.cr.usgs.gov/srtm/version2_1/Documentation/Quickstart.pdf) it says data voids are assigned the value -32768. Should I be using the NASA value for -srcnodata and -dstnodata instead? Cheers Adam Ratcliffe Software Developer GeoSmart PO Box 101001 29 Apollo Drive, Mairangi Bay, Auckland, NZ Ph: +64 9 9668757 Fx: +64 9 966 8721 Email: adam.ratcliffe@... Web: www.GeoSmart.co.nz SmartNAV - The premier Car Navigation dataset for NZ SmartFIND - Web and Mobile mapping solutions by GeoSmart www.aamaps.co.nz - For FREE web maps and directions within NZ Route2GO - Optimised journey planning to create the most efficient routes to your customers for your sales or service calls _______________________________________________ Mapnik-users mailing list Mapnik-users@... https://lists.berlios.de/mailman/listinfo/mapnik-users |
|
|
Re: Black holes in hillshade image
by Adam Ratcliffe-2
::
Rate this Message:
With further experimentation it seems that 32767 does seem to be the correct value for the -srcnodata option. I've also found that if I omit the -dstnodata option from the gdalwarp command the black holes do not appear, presumably because the void data is then not included in the reprojected image. Cheers Adam On 1/10/2009, at 3:56 PM, Adam Ratcliffe wrote:
Adam Ratcliffe Software Developer GeoSmart PO Box 101001 29 Apollo Drive, Mairangi Bay, Auckland, NZ Ph: +64 9 9668757 Fx: +64 9 966 8721 Email: adam.ratcliffe@... Web: www.GeoSmart.co.nz SmartNAV - The premier Car Navigation dataset for NZ SmartFIND - Web and Mobile mapping solutions by GeoSmart www.aamaps.co.nz - For FREE web maps and directions within NZ Route2GO - Optimised journey planning to create the most efficient routes to your customers for your sales or service calls _______________________________________________ Mapnik-users mailing list Mapnik-users@... https://lists.berlios.de/mailman/listinfo/mapnik-users |
|
|
Re: Black holes in hillshade image
by Holger Schöner
::
Rate this Message:
Hi Adam,
On Thu, 1 Oct 2009, Adam Ratcliffe wrote: > With further experimentation it seems that 32767 does seem to be the correct > value for the -srcnodata option. I've also found that if I omit the > -dstnodata option from the gdalwarp command the black holes do not appear, > presumably because the void data is then not included in the reprojected > image. > On 1/10/2009, at 3:56 PM, Adam Ratcliffe wrote: >> I've created a hillshade image using the NASA SRTM3 data following the >> notes at: http://wiki.openstreetmap.org/wiki/HikingBikingMaps. Looks great >> except for the occasional 'black hole' in the image, I expect this is due >> to data voids not being rendered correctly. >> >> What's got me puzzled about this is that the wiki page suggests using a >> value of 32767 for the -srcnodata and -dstnodata options to gdalwarp, >> howver in the NASA docs >> (http://dds.cr.usgs.gov/srtm/version2_1/Documentation/Quickstart.pdf) it >> says data voids are assigned the value -32768. >> >> Should I be using the NASA value for -srcnodata and -dstnodata instead? I wrote that Wiki page, and I just remember that at that time (a year ago) the interpretation was unclear for me. I remember having some problems and experimenting; but in my region, there are no voids, so I might not have found the right solution at that time. And I remember encountering "black holes" later on in other regions (thinking there is no way around having them, without setting them explicitly to some arbitrary value), as well. I will try out your suggestions in a few days myself (and might update the page afterwards, if you have not done so then) ... Thanks, Holger _______________________________________________ Mapnik-users mailing list Mapnik-users@... https://lists.berlios.de/mailman/listinfo/mapnik-users _______________________________________________ Mapnik-users mailing list Mapnik-users@... https://lists.berlios.de/mailman/listinfo/mapnik-users |
|
|
Re: Black holes in hillshade image
by Adam Ratcliffe-2
::
Rate this Message:
Hi Holger,
I've edited the wiki page, just removed the -dstnodata option to gdalwarp - that works for me! Cheers Adam ________________________________________ From: Holger Schöner [numenor@...] Sent: 02 October 2009 22:09 To: Adam Ratcliffe Cc: mapnik-users@... Subject: Re: [Mapnik-users] Black holes in hillshade image Hi Adam, On Thu, 1 Oct 2009, Adam Ratcliffe wrote: > With further experimentation it seems that 32767 does seem to be the correct > value for the -srcnodata option. I've also found that if I omit the > -dstnodata option from the gdalwarp command the black holes do not appear, > presumably because the void data is then not included in the reprojected > image. > On 1/10/2009, at 3:56 PM, Adam Ratcliffe wrote: >> I've created a hillshade image using the NASA SRTM3 data following the >> notes at: http://wiki.openstreetmap.org/wiki/HikingBikingMaps. Looks great >> except for the occasional 'black hole' in the image, I expect this is due >> to data voids not being rendered correctly. >> >> What's got me puzzled about this is that the wiki page suggests using a >> value of 32767 for the -srcnodata and -dstnodata options to gdalwarp, >> howver in the NASA docs >> (http://dds.cr.usgs.gov/srtm/version2_1/Documentation/Quickstart.pdf) it >> says data voids are assigned the value -32768. >> >> Should I be using the NASA value for -srcnodata and -dstnodata instead? Thanks for working this out! Maybe you can correct the wiki page? I wrote that Wiki page, and I just remember that at that time (a year ago) the interpretation was unclear for me. I remember having some problems and experimenting; but in my region, there are no voids, so I might not have found the right solution at that time. And I remember encountering "black holes" later on in other regions (thinking there is no way around having them, without setting them explicitly to some arbitrary value), as well. I will try out your suggestions in a few days myself (and might update the page afterwards, if you have not done so then) ... Thanks, Holger _______________________________________________ Mapnik-users mailing list Mapnik-users@... https://lists.berlios.de/mailman/listinfo/mapnik-users |
|
|
Re: Black holes in hillshade image
by Holger Schöner
::
Rate this Message:
Am Fr, 2. Oktober 2009 schrieb Holger Schöner:
> On Thu, 1 Oct 2009, Adam Ratcliffe wrote: > > With further experimentation it seems that 32767 does seem to be the > > correct value for the -srcnodata option. I've also found that if I > > omit the -dstnodata option from the gdalwarp command the black holes do > > not appear, presumably because the void data is then not included in > > the reprojected image. > > Thanks for working this out! Maybe you can correct the wiki page? Thanks for improving it! > I will try out your suggestions in a few days myself (and might update > the page afterwards, if you have not done so then) ... I tried it out, and I can confirm that the void handling is improved by leaving out the -dstnodata option. Of course, artefacts are still to be seen (at least in hill-shaded and relief-colored results, that is what I tested), but they are less striking. -- Holger Schoener numenor@... _______________________________________________ Mapnik-users mailing list Mapnik-users@... https://lists.berlios.de/mailman/listinfo/mapnik-users |
| Free embeddable forum powered by Nabble | Forum Help |
