as promised, this is the white paper in which I attempt to
outline a unified framework for implementing geophysical
functionality in GRASS.
What is missing:
- anything relating to 3D analysis
- color ramp modification tools
- a full list of the actual filter functionality we want to
implement (see section at bottom of document),
should also include a short description of how each filter
operates and what it is useful for
I am attaching a PDF version for viewing and a M$Word version
for your edits.
I would like everyone interested to add comments, make changes
and additions, look for inconsistencies (terminology), spelling,
grammar (I am not a native English speaker), clarifications
- please mark your changes with another colour in the DOC file
and send it back to me, so I can synchronise all the edits.
Craig would then be able to forward the white paper for review.
Do not forget to put your names on the authors list.
Let me know what you think about this. The framework so far
mainly reflects my personal work experience and ideas.
Benjamin.
(attachments)
white_paper.25858DEFANGED-doc (48.5 KB)
white_paper.pdf (303 KB)
Hi Martin,
I think we should wait a bit longer until the geophysical
modules have actually been written and there has been
some feedback on the issue, as well.
Once we get to that stage, we should be able to produce
a very nice contribution to the GRASS newsletter, which
without a doubt, would find a lot of interested readers.
But I feel a bit uncomfortable about stirring up very high
hopes and not being able to deliver ...
I have another suggestion for a contribution I could make:
I have written a set of modules for GRASS 5.3 that allow
Belief-based spatial prediction.
It uses Dempster-Shafer's Theory of Evidence which is
sort of a generalised probability theory.
My work is very similar to something that was done for
ArcGIS a while back (see:
http://gis.esri.com/library/userconf/proc99/proceed/papers/pap295/p295.htm
... but of course, being GRASS-based it's much more
flexible and powerful 
I used the system very successfully to predict locations
of archaeological sites for heritage management in
North-eastern Germany.
You can find a summary in:
http://www.caa2004.org/abstracts/kamermans.doc
(search for "Benjamin Ducke" in the document to find my abstract)
Let me know if this is a sort of topic that would be
suited for the GRASS newsletter and I will write up something.
Regards,
Benjamin
On Sat, 29 May 2004 18:45:58 +0200
Martin Wegmann <m.wegmann@web.de> wrote:
Hello Benjamin,
I followed your discussion on the list quit a bit although I am not (yet)
interested in this kind of modules, but I would welcome if you like to extend
your white paper to an article for the GRASSNewsletter.
Via this way you can elaborate your ideas, add some pictures and announce what
you are planning to do.
If you are interested ,please have a look at
http://grass.itc.it/newsletter/index
best regards, Martin
Hello Benjamin,
On Saturday 29 May 2004 22:46, Benjamin Ducke wrote:
[...]
I have another suggestion for a contribution I could make:
I have written a set of modules for GRASS 5.3 that allow
Belief-based spatial prediction.
It uses Dempster-Shafer's Theory of Evidence which is
sort of a generalised probability theory.
it sounds pretty interesting especially because Fabien Fivaz will write an
article featuring non-linear spatial modelling via GRASP-R
Let me know if this is a sort of topic that would be
suited for the GRASS newsletter and I will write up something.
[...]
yes, I think every kind of GRASS capabilities will be suited for the GRASS
Newsletter - there does a wide variety of GRASS user exist and everyone uses
a different set of capabilities -- I am myself still astonished which
capabilities can be found inside GRASS.
When your geophysical module work reached a certain stage (that you can be
sure to deliver all hopes ), it would be great to read something
covering this topic.
regards, Martin
On Sat, 29 May 2004 18:45:58 +0200
Martin Wegmann <m.wegmann@web.de> wrote:
> Hello Benjamin,
>
> I followed your discussion on the list quit a bit although I am not (yet)
> interested in this kind of modules, but I would welcome if you like to
> extend your white paper to an article for the GRASSNewsletter.
> Via this way you can elaborate your ideas, add some pictures and announce
> what you are planning to do.
> If you are interested ,please have a look at
> http://grass.itc.it/newsletter/index
>
> best regards, Martin