I have a number of rasters from white I need to extract building polygons. However, the rasters are in black and white, and buildings, land and wood will be symbolized differently by different patterns.
How could I identify the different patterns and filter the polygons in different streams depending on their pattern?
It's not really difficult to get polygons out of raster cells, but
interpreting patterns is the challenge here.
How distinct are the patterns? Could you send a screenshot?
> I have a number of rasters from white I need to extract building polygons.
> However, the rasters are in black and white, and buildings, land and wood
> will be symbolized differently by different patterns.
> How could I identify the different patterns and filter the polygons in
> different streams depending on their pattern?
On Friday, June 29, 2012 2:33:37 PM UTC+1, Klaas Dijkstra wrote:
> Hi Maryse,
> It's not really difficult to get polygons out of raster cells, but > interpreting patterns is the challenge here. > How distinct are the patterns? Could you send a screenshot?
>> I have a number of rasters from white I need to extract building >> polygons. However, the rasters are in black and white, and buildings, land >> and wood will be symbolized differently by different patterns.
>> How could I identify the different patterns and filter the polygons in >> different streams depending on their pattern?
>> You received this message because you are subscribed to the Google Groups >> "FME Talk" group.
>> To post to this group, send email to fmetalk@googlegroups.com >> To unsubscribe from this group, send email to >> fmetalk-unsubscribe@googlegroups.com
Hi Maryse,
There's a RasterToPolygonCoercer transformer in FME2012, but I really
don't hold out much hope. Those images are going to be really tough to
turn into vector features.
The only thought I have is to turn it into an 8-bit greyscale (I'm
assuming black+white = 1bit image?) and resample it to a coarser
resolution. That might help pick out the shading of the buildings, but
I'm not certain.
This page on fmepedia also shows vectorization with FME, but again I'm
not sure if it will work with your images:
http://fmepedia.safe.com/articles/Samples_and_Demos/Using-Conditions-...
Sorry to be negative. I just don't want you to waste too much time in
a losing battle. I fear that FME is not going to be the best tool for
this job. But if you find a way then please let us know - I'd be
interested to see your method.
Regards
Mark
Mark Ireland | Senior Product Specialist
Safe Software Inc.
Suite 2017, 7445-132nd Street, Surrey, BC, CANADA
T 604.501.9985 | F 604.501.9965
http://www.safe.com/support | Twitter @FMEDoctors | http://www.fmepedia.com
On Jun 29, 7:01 am, Maryse Fantin <maryse.fan...@googlemail.com>
wrote:
> On Friday, June 29, 2012 2:33:37 PM UTC+1, Klaas Dijkstra wrote:
> > Hi Maryse,
> > It's not really difficult to get polygons out of raster cells, but
> > interpreting patterns is the challenge here.
> > How distinct are the patterns? Could you send a screenshot?
> >> I have a number of rasters from white I need to extract building
> >> polygons. However, the rasters are in black and white, and buildings, land
> >> and wood will be symbolized differently by different patterns.
> >> How could I identify the different patterns and filter the polygons in
> >> different streams depending on their pattern?
> >> You received this message because you are subscribed to the Google Groups
> >> "FME Talk" group.
> >> To post to this group, send email to fmetalk@googlegroups.com
> >> To unsubscribe from this group, send email to
> >> fmetalk-unsubscribe@googlegroups.com
I agree with Mark, that this is a challenge!
Especially because the pictures are not of great quality.
I worked with b/w raster, but only to remove noise from them (by
turning them into polygons, identifying/removing the smaller black
bits and then turning them into raster again).
Klaas Dijkstra
On 30 jun, 00:17, Mark Ireland <mark.irel...@safe.com> wrote:
> Hi Maryse,
> There's a RasterToPolygonCoercer transformer in FME2012, but I really
> don't hold out much hope. Those images are going to be really tough to
> turn into vector features.
> The only thought I have is to turn it into an 8-bit greyscale (I'm
> assuming black+white = 1bit image?) and resample it to a coarser
> resolution. That might help pick out the shading of the buildings, but
> I'm not certain.
> This page on fmepedia also shows vectorization with FME, but again I'm
> not sure if it will work with your images:http://fmepedia.safe.com/articles/Samples_and_Demos/Using-Conditions-...
> Sorry to be negative. I just don't want you to waste too much time in
> a losing battle. I fear that FME is not going to be the best tool for
> this job. But if you find a way then please let us know - I'd be
> interested to see your method.
> Regards
> Mark
> Mark Ireland | Senior Product Specialist
> Safe Software Inc.
> Suite 2017, 7445-132nd Street, Surrey, BC, CANADA
> T 604.501.9985 | F 604.501.9965http://www.safe.com/support| Twitter @FMEDoctors |http://www.fmepedia.com
> On Jun 29, 7:01 am, Maryse Fantin <maryse.fan...@googlemail.com>
> wrote:
> > Here is a screenshot of 2 of them.
> > Let me know if you have any idea. I'm thinking about calculating the
> > density of black points withing the polygons.
> > On Friday, June 29, 2012 2:33:37 PM UTC+1, Klaas Dijkstra wrote:
> > > Hi Maryse,
> > > It's not really difficult to get polygons out of raster cells, but
> > > interpreting patterns is the challenge here.
> > > How distinct are the patterns? Could you send a screenshot?
> > >> I have a number of rasters from white I need to extract building
> > >> polygons. However, the rasters are in black and white, and buildings, land
> > >> and wood will be symbolized differently by different patterns.
> > >> How could I identify the different patterns and filter the polygons in
> > >> different streams depending on their pattern?
> > >> You received this message because you are subscribed to the Google Groups
> > >> "FME Talk" group.
> > >> To post to this group, send email to fmetalk@googlegroups.com
> > >> To unsubscribe from this group, send email to
> > >> fmetalk-unsubscribe@googlegroups.com
