Dear list,
I produced an uncontinuous distance map as a result of r.grow.distance. Most
parts of the distance map are smooth but there are some sharp edges in the
map. This means that with a 100m resolution of the map in most parts of the
map the distance value increases about 100 m with every pixel. At the sharp
edges there are leaps of up to 1000m from pixel to pixel. Depending on the
color map used for displaying the distance map you can also see these jumps
clearly.
Has anyone experienced the same?
Cheers Falko
--
View this message in context: http://osgeo-org.1803224.n2.nabble.com/r-grow-distance-tp5100098p5100098.html
Sent from the Grass - Users mailing list archive at Nabble.com.
fengel wrote:
I produced an uncontinuous distance map as a result of r.grow.distance. Most
parts of the distance map are smooth but there are some sharp edges in the
map. This means that with a 100m resolution of the map in most parts of the
map the distance value increases about 100 m with every pixel. At the sharp
edges there are leaps of up to 1000m from pixel to pixel. Depending on the
color map used for displaying the distance map you can also see these jumps
clearly.
Can you provide a test case?
--
Glynn Clements <glynn@gclements.plus.com>
Here ist the test case. I also added my result of r.grow.distance.
Falko
http://osgeo-org.1803224.n2.nabble.com/file/n5102335/test_case.zip
test_case.zip
--
View this message in context: http://osgeo-org.1803224.n2.nabble.com/r-grow-distance-tp5100098p5102335.html
Sent from the Grass - Users mailing list archive at Nabble.com.
fengel wrote:
Here ist the test case. I also added my result of r.grow.distance.
That contains two maps: test_bool is empty (all cells are -32768),
while test_dist appears to be the output.
--
Glynn Clements <glynn@gclements.plus.com>
Sorry for that. I hope this one's ok...
http://osgeo-org.1803224.n2.nabble.com/file/n5107306/test_case.zip
test_case.zip
--
View this message in context: http://osgeo-org.1803224.n2.nabble.com/r-grow-distance-tp5100098p5107306.html
Sent from the Grass - Users mailing list archive at Nabble.com.
fine in 7.0, broken in 6.x, this fixes it for me in 6.5:
Index: main.c
--- main.c (revision 42353)
+++ main.c (working copy)
@@ -287,6 +287,12 @@
check(col, 1, -1);
}
+ for (col = 0; col < ncols; col++)
+ check(col, -1, 0);
+
+ for (col = ncols - 1; col >= 0; col--)
+ check(col, 1, 0);
+
if (dist_name) {
if (out_row != dist_row)
for (col = 0; col < ncols; col++)
makes sense?
Markus M
fengel wrote:
Sorry for that. I hope this one's ok...
http://osgeo-org.1803224.n2.nabble.com/file/n5107306/test_case.zip
test_case.zip
--
View this message in context: http://osgeo-org.1803224.n2.nabble.com/r-grow-distance-tp5100098p5107306.html
Sent from the Grass - Users mailing list archive at Nabble.com.
_______________________________________________
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user
Markus Metz wrote:
fine in 7.0, broken in 6.x, this fixes it for me in 6.5:
Right; this was done for 7.0 in r34429, but it was mixed up with
the addition of metric=geodesic, so it got overlooked.
--
Glynn Clements <glynn@gclements.plus.com>
Hi!
Sorry for taking so long.
I am not familiar with c and also I didn't really get the idea of your fix.
So there is something wrong with the total numbers of columns that
r.grow.distance counts?
Do you use your code as a kind of patch? I am using grass 6.4. I am also not
familiar with patching things. Is there a way for me do use your fix?
Falko
--
View this message in context: http://osgeo-org.1803224.n2.nabble.com/r-grow-distance-tp5100098p5153768.html
Sent from the Grass - Users mailing list archive at Nabble.com.
fengel wrote:
I am not familiar with c and also I didn't really get the idea of your fix.
So there is something wrong with the total numbers of columns that
r.grow.distance counts?
Do you use your code as a kind of patch? I am using grass 6.4. I am also not
familiar with patching things. Is there a way for me do use your fix?
The change has been committed to the 6.4 and 6.5 branches:
6.4: http://trac.osgeo.org/grass/changeset/42412
6.5: http://trac.osgeo.org/grass/changeset/42413
If you build from source, update to the latest SVN revision.
Or you can try the latest binary snapshot:
http://grass.osgeo.org/grass64/binary/linux/snapshot/
--
Glynn Clements <glynn@gclements.plus.com>