HI Umit, The --terrain option enables the use osgTerrain::TerrainTile rather the use of osg::Geometry based tiles, and osgTerrain supports both geocentric and flat earth projections.
Robert. On Fri, Dec 12, 2008 at 8:37 PM, Ümit Uzun <[email protected]> wrote: > Hi Robert, > > Thanks for clarification. I have thought --geocentric options generally > using for spherical terrain like whole earth terrain database (i.e > bluemarble world) and --terrain option for flat terrain. And have used both > option seperately to create terrain but haven't tried both of them. So I > misunderstood usage. > > Regards. > > 2008/12/12 christophe loustaunau <[email protected]> >> >> Hi, >> >> I have test your data and there is strange things. >> If I use this command line : >> osgdem --terrain --geocentric -t hillsmall -o hillsmall.ive >> osgdem reproject the two Geotiff and create two temporary files. But it >> seems not to use it, as in the output database >> the tiles no not line up. >> >> But if you put these two temporary files into the directory /hillsmall and >> run again osgdem with the same command line >> osgdem --terrain --geocentric -t hillsmall -o hillsmall.ive >> The output database is ok, everything lines up ! (see the sceenshot). >> >> I think osgdem do not use the temporary files, but I don't know why. Maybe >> you could debug it. >> Anyway, you could reproject you files and then everything will work fine ! >> >> Hope this help. >> >> Regards. >> >> On Fri, Dec 12, 2008 at 7:46 PM, Tueller, Shayne R Civ USAF AFMC 519 >> SMXS/MXDEC <[email protected]> wrote: >>> >>> All, >>> >>> I finally have two small Geotiff tiles (1m resolution) that I've >>> experimented with to examine this problem. I run them through VPB and >>> they >>> first get reprojected (via GDAL I presume) into temporary files. I then >>> load >>> the two temporary reprojected files into OpenEV for examination and they >>> still appear to line up as they did before being reprojected. That would >>> seem to suggest that the problem isn't with the reprojection step (or >>> what >>> GDAL is doing). >>> >>> Once VPB installs them into the database, they do NOT line up as I have >>> shown in my previous email with the screen capture. This would suggest >>> that >>> VPB is doing something that is shifting them off so they don't align. >>> >>> I've attached the two GeoTiff tiles if anyone wants to reproduce this >>> problem. >>> >>> Using VPB, I'm issuing the command: >>> >>> osgdem --TERRAIN --PagedLOD --geocentric -t hillsmall -l 8 -o >>> hillsmall.ive >>> >>> The "hillsmall" directory contains the two tiles. Interestingly, when I >>> remove the "--geocentric" flag from the above command, things look fine >>> for >>> a flat database and the tiles align (which isn't surprising). It's almost >>> as >>> if the spheroid between the tiles and what VPB is using are not matched. >>> >>> Any ideas on what's going on? >>> >>> -Shayne >>> >>> -----Original Message----- >>> From: [email protected] >>> [mailto:[email protected]] On Behalf Of J.P. >>> Delport >>> Sent: Wednesday, December 10, 2008 11:39 PM >>> To: OpenSceneGraph Users >>> Subject: Re: [osg-users] discontinuity in texture in VPB... >>> >>> Hi, >>> >>> I'd also suggest what Christophe said. Do the projection of some tiles >>> manually before VPB is run. I normally use something like: >>> >>> gdalwarp -t_srs "+proj=latlong +datum=WGS84" -r bilinear $name >>> ../reprojected/$newname >>> >>> Then load the tiles into something like QGIS (http://www.qgis.org/) to >>> see if they line up. >>> >>> jp >>> >>> christophe loustaunau wrote: >>> > Hi Shayne, >>> > >>> > Maybe you could reproject your imagery to a geographic projection with >>> > gdal (gdal_translate) and see if you have the same artifact. >>> > You will know if it's a VPB or a gdal reprojection problem. >>> > >>> > On Wed, Dec 10, 2008 at 6:14 PM, Tueller, Shayne R Civ USAF AFMC 519 >>> > SMXS/MXDEC <[email protected] >>> > <mailto:[email protected]>> wrote: >>> > >>> > I've built a geocentric database using VBP 0.9.7. <http://0.9.7.> >>> > The imagery is GeoTiff tiles that are NAIP 1 meter UTM and the >>> > elevation is DTED level 1. The GeoTiffs appear to be reprojected by >>> > VPB (I assume using GDAL internally) for the geocentric database >>> > setting. Things appear to build just fine without error but when I >>> > view the database in OSG, I see a nasty discontinuity across >>> > texture >>> > tile boundaries. You can clearly see this in the airfield picture >>> > that is attached. The landing strip does not line up across the >>> > tile >>> > boundary. I've verified that the tiles themselves are correct >>> > across >>> > boundaries before I fed them into VPB. The discontinuity appears >>> > regardless of the LOD that is being paged in. >>> > >>> > >>> > >>> > My question is, is this an artifact that I must live with since the >>> > original imagery is based in UTM and the reprojection that VPB does >>> > (or GDAL) introduces inaccuracies when going to a geocentric >>> > database? Is there something else I can do to eliminate these >>> > discontinuities? >>> > >>> > >>> > >>> > Any input or ideas anyone can suggest to eliminate these >>> > discontinuities in the database would be welcome. >>> > >>> > >>> > >>> > Thanks, >>> > >>> > -Shayne >>> > >>> > >>> > _______________________________________________ >>> > osg-users mailing list >>> > [email protected] >>> > <mailto:[email protected]> >>> > >>> http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org >>> > >>> > >>> > >>> > >>> > -- >>> > Christophe Loustaunau. >>> > >>> > >>> > >>> > ------------------------------------------------------------------------ >>> > >>> > _______________________________________________ >>> > osg-users mailing list >>> > [email protected] >>> > >>> > http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org >>> >>> -- >>> This message is subject to the CSIR's copyright terms and conditions, >>> e-mail >>> legal notice, and implemented Open Document Format (ODF) standard. >>> The full disclaimer details can be found at >>> http://www.csir.co.za/disclaimer.html. >>> >>> This message has been scanned for viruses and dangerous content by >>> MailScanner, >>> and is believed to be clean. MailScanner thanks Transtec Computers for >>> their support. >>> >>> _______________________________________________ >>> osg-users mailing list >>> [email protected] >>> http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org >>> >>> _______________________________________________ >>> osg-users mailing list >>> [email protected] >>> http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org >>> >> >> >> >> -- >> Christophe Loustaunau. >> >> _______________________________________________ >> osg-users mailing list >> [email protected] >> http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org >> > > > > -- > Ümit Uzun > > _______________________________________________ > osg-users mailing list > [email protected] > http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org > > _______________________________________________ osg-users mailing list [email protected] http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
