[GRASS-user] Tessellation of a set of polygons

Tue Feb 25 16:21:01 PST 2014

I've used Theissen/Voroni polygons but tessellation is new to me, not surprising given my state of knowledge so I would appreciate how it differs from T/V polygons for my edification.

Thanks in advance,

Lyle Browning

On Feb 25, 2014, at 11:54 AM, Thomas Leduc <thomas.leduc at cerma.archi.fr> wrote:

> Dear GRASS-users & developpers,
> 
> As summed up below, I've tried to tessellate the enclosed datasource (a shapefile composed of several building footprints) using 2 different approaches:
> 
> 	g.remove vect=buildings,plots rast=buildings_r,plots_r
> 	v.in.ogr -o dsn=/tmp/buildings.shp output=buildings
> 
> 1st method:
> 	v.voronoi -a --overwrite input=buildings output=vor smoothness=0.15
> 	v.out.ogr --overwrite input=vor type=area dsn=/tmp/src olayer=vor layer=1 format=ESRI_Shapefile
> 
> 2nd method:
> 	g.region vect=buildings nsres=0.25 ewres=0.25
> 	v.to.rast --overwrite input=buildings output=buildings_r use=attr attrcolumn=gid
> 	r.grow --overwrite input=buildings_r output=plots_r metric=euclidean radius=4800
> 	r.to.vect --o input=plots_r output=plots type=area
> 	v.out.ogr --o input=plots type=area dsn=/tmp/src olayer=plots layer=1 format=ESRI_Shapefile
> 
> As may be observed with the screenshot, (both) output results does not match expectations.
> 
> Do you have any idea of the reason why it's wrong?
> 
> Many thanks for your help,
> 
> Thomas
> 
> <tessellation.png>
> 
> 
> 
> On Tue, Sep 10, 2013 at 3:10 PM, Thomas Leduc <thomas.leduc at cerma.archi.fr> wrote:
> Dear Markus,
> 
> Impressive responsiveness. Thanks for the dedicated development. http://trac.osgeo.org/grass/changeset/57621 is fully operative on the given dataset.
> 
> It took me some time to test it, because I'm a complete newby with GRASS community. Just to keep it written somewhere, I first had to checkout, configure, compile, and install GRASS7 snapshot on my Debian 6.0.7:
> 
>     svn checkout https://svn.osgeo.org/grass/grass/trunk grass7_trunk
>     cd ./grass7_trunk
>     ./configure --enable-64bit -with-freetype-includes=/usr/include/freetype2 --prefix=/usr/local/misc/grass7-snapshot --with-sqlite
>     make
>     make install
> 
> Then, in the GRASS console, I've copied and pasted the following instructions:
> 
>     g.remove vect=buildings,vor
>     v.in.ogr -o dsn=/tmp/src/src.shp output=buildings
>     v.voronoi -a --overwrite input=buildings output=vor segment=0.15
>     v.out.ogr --overwrite input=vor type=area dsn=/tmp/src olayer=vor layer=1 format=ESRI_Shapefile
> 
> Many thanks once again.
> 
> Thomas
> 
> 
> On Sun, Sep 8, 2013 at 7:49 PM, Markus Metz <markus.metz.giswork at gmail.com> wrote:
> On Thu, Aug 29, 2013 at 5:38 PM, Thomas Leduc
> <thomas.leduc at cerma.archi.fr> wrote:
> > Let's consider a set of input polygons (represented by dark gray polygonal
> > footprints in the enclosed screenshot [1]). I wonder whether there already
> > exists a (set of) GRASS instructions to produce the "surrounding
> > tessellation". I mean, could you please tell me how to compute the set of
> > multicolor polygons in the enclosed screenshot?
> >
> > As an example, the blue polygon is the set of points that are:
> >   - overlayed by the dark gray polygon labeled #1
> >   - or closer to the dark gray polygon labeled #1 than to any other input
> > polygon.
> >
> I have added tessellation of areas to v.voronoi in trunk r57621. You
> need to use the new -a flag and can control the smoothness of the
> output with the new segment option. With your example, segment=0.15
> produces reasonably smooth output.
> 
> This implementation ignores shared boundaries (not present in your
> example) which would need to be patched in afterwards. An alternative
> would be to use v.overlay to obtain areas outside the original input
> areas.
> 
> HTH,
> 
> Markus M
> 
> > Thanks a lot for any help/advice.
> >
> > Thomas Leduc
> >
> > PS: I'm aware there already exist in GRASS some functions such as
> > "v.to.rast, r.grow.distance, and r.contour" which could help in building map
> > of distances, and isocontours... I'm also aware there already exist
> > "v.to.rast, r.thin, and r.to.vect feature=line" to build some sort of medial
> > axis transform or skeleton.
> >
> > PPS: input data set
> >
> > create table b(gid int, the_geom GEOMETRY(POLYGON, 2D));
> > insert into b values (0, GeomFromText('POLYGON ((0 0, 10 0, 10 10, 0 10, 0
> > 0))'));
> > insert into b values (1, GeomFromText('POLYGON ((30 0, 30 -30, 10 -30, 10
> > -40, 40 -40, 40 0, 30 0))'));
> > insert into b values (2, GeomFromText('POLYGON ((20 20, 100 20, 100 40, 60
> > 40, 60 30, 40 30, 40 60, 20 60, 20 20))'));
> > insert into b values (3, GeomFromText('POLYGON ((60 10, 60 -20, 100 -10, 100
> > 10, 60 10))'));
> > insert into b values (4, GeomFromText('POLYGON ((-10 -10, -10 -40, -40 -40,
> > -40 -20, -60 -20, -60 0, -50 0, -50 -10, -10 -10))'));
> > insert into b values (5, GeomFromText('POLYGON ((-40 20, -40 40, -10 40, -10
> > 30, -20 20, -40 20))'));
> >
> > [1] this screenshot of about 11 KB is also downloadable at
> > https://dl.dropboxusercontent.com/u/8846569/tessellation.png
> >
> >
> > _______________________________________________
> > grass-user mailing list
> > grass-user at lists.osgeo.org
> > http://lists.osgeo.org/mailman/listinfo/grass-user
> _______________________________________________
> grass-user mailing list
> grass-user at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-user
> 
> 
> 
> -- 
> Thomas LEDUC
> 
> 
> 
> -- 
> Thomas LEDUC
> <buildings.zip>_______________________________________________
> grass-user mailing list
> grass-user at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-user

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-user/attachments/20140225/d0a96200/attachment.html>