I'm having a problem trying to execute a buffer of 150m on a polygon
generated from a raster with resolution 5x5.
The polygon is a single area, constituted of 1576 vertices (highly granular).
v.buffer generates 32633 vertices, the it get "mad" building the
topology: it tries to remove intersections, duplicates, etc. giving
the following final result:
Topology was built.
Number of nodes : 32633
Number of primitives: 65264
Number of points : 0
Number of lines : 0
Number of boundaries: 65264
Number of centroids : 0
Number of areas : 32632
Number of isles : 1
Number of areas without centroid : 32632
The vector is not readable...
I don.t know if the problems is due to the buffer algorithm, which
fails because of the pixelated polygon (figure below)
I've tried to use a *convex* polygon with the same characteristics
(built from 5x5 raster). The buffer has worked right.
So, I suppose it depends on the non-convexity on the previous (as I
show in the ascii-art figure).
I'm having a problem trying to execute a buffer of 150m on a polygon
generated from a raster with resolution 5x5.
The polygon is a single area, constituted of 1576 vertices (highly
granular).
v.buffer generates 32633 vertices, the it get "mad" building the
topology: it tries to remove intersections, duplicates, etc. giving
the following final result:
Topology was built.
Number of nodes : 32633
Number of primitives: 65264
Number of points : 0
Number of lines : 0
Number of boundaries: 65264
Number of centroids : 0
Number of areas : 32632
Number of isles : 1
Number of areas without centroid : 32632
The vector is not readable...
I don.t know if the problems is due to the buffer algorithm, which
fails because of the pixelated polygon (figure below)
It's not a perfect solution but maybe try v.to.rast + r.buffer +
r.to.vect ?
Hamish
In general I try to avoid rasterising of vectors, do something and then
vectorise back. It's off topic but I have a question: what's the deal
concerning loss off information/ distortion etc. ? On which factors it
depends most? Resolution?
Sorry for this interruption and thank you in advance,
Hamish wrote:
> v.buffer has some problems,
> http://trac.osgeo.org/grass/ticket/90
>
> It's not a perfect solution but maybe try v.to.rast + r.buffer +
> r.to.vect ?
Nikos:
In general I try to avoid rasterising of vectors, do something and
then vectorise back.
it's not a great solution, but it is perhaps better than having output
with incorrect gaps.
It's off topic but I have a question: what's
the deal concerning loss off information/ distortion etc. ? On which
factors it depends most? Resolution?
resolution + rasterization method.
you can minimize resolution issues by making the resolution as fine as
practicable; see the v.to.rast module help page for information on the
different methods depending on feature type.
Hamish
____________________________________________________________________________________
Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs
> Hamish wrote:
> > v.buffer has some problems,
> > http://trac.osgeo.org/grass/ticket/90
> >
> > It's not a perfect solution but maybe try v.to.rast + r.buffer +
> > r.to.vect ?
Nikos:
> In general I try to avoid rasterising of vectors, do something and
> then vectorise back.
it's not a great solution, but it is perhaps better than having output
with incorrect gaps.
> It's off topic but I have a question: what's
> the deal concerning loss off information/ distortion etc. ? On which
> factors it depends most? Resolution?
resolution + rasterization method.
you can minimize resolution issues by making the resolution as fine as
practicable; see the v.to.rast module help page for information on the
different methods depending on feature type.
This is probably bit late, but your problem is in type=area. Try
type=boundary or type=line or any combination of those or, actually,
ommit the type parameter at all.... Or if it does not help, play a bit
with reduction and/or threshold parameter.