[postgis-users] RE : How does ST_Slope works ?

[Francois Hugues]

Hello,

Thanks for your answer.

If some functions can't handle all the pixels, this is a big limit for the use of tiled rasters if some function can't handle all the pixels although tiling and indexing raster is a major improvement (thanks to postgis).It's going to be tricky if we need to have both tiled and untiled version for the same raster.

Hugues.


-------- Message d'origine--------
De: postgis-users-bounces at postgis.refractions.net de la part de Pierre Racine
Date: mer. 13/06/2012 21:13
À: PostGIS Users Discussion
Objet : Re: [postgis-users] How does ST_Slope works ?
 
> 1. To know which Pixeltype I had to use, I tried to use st_bandpixeltype which
> told me my pixels were 32BSI type but my results using this type are 0 or 1 and if
> I try another type like 32BF, results seems to be more useful. Does someone
> have an explanation about this weird difference ? What is the unit used by
> st_slope ?

The results are in radian so rounded to 0 or 1 when the pixel type is integer.

> 2. My raster is a tiled one and as you can see on both pictures, ST_Slope gives no
> results for pixels located at the border of each tile. I can understand it is
> impossible to find the slope for the pixels which make the border of the raster,
> but is there not a way to have complete results for the tile in the middle of the
> raster and thus use the neighboring tiles.

Not yet. The only way is to load the raster untiled or to ST_Union them which might be slow.

Pierre
_______________________________________________
postgis-users mailing list
postgis-users at postgis.refractions.net
http://postgis.refractions.net/mailman/listinfo/postgis-users

-------------- next part --------------
A non-text attachment was scrubbed...
Name: winmail.dat
Type: application/ms-tnef
Size: 3639 bytes
Desc: not available
URL: <http://lists.osgeo.org/pipermail/postgis-users/attachments/20120614/5c41754a/attachment.bin>