[GRASS-user] v.generalize / reanimation of dead lines ?

[Benjamin Ducke]

Further on this subject, I tried v.generalize on the North Carolina
sample dataset and ran into the same "attempt to read dead line" error
with the "boundary_county" areal layer. Very small threshold settings
( < 1.0 ) worked OK, but obviously, there is only a minimal
simplification to be gained from this. Any threshold setting around 10.0
resulted in the above error not more than 10% through the data.
I tried all different algorithms and did not have better success with
any of them.

Best,

Ben

Markus Metz wrote:
> Hi Peter,
> 
> as an interim solution you might try an alternative to v.generalize that
> I called v.simplify, available here:
> http://markus.metz.giswork.googlepages.com/line_simplification.tar.gz
> The module works for me so far, but I still discover strange behaviour
> now and then. I developed that module specifically for areas, before
> v.generalize was available.
> Differences to v.generalize are that this module only
> simplifies/generalizes lines/boundaries, smoothing is not available, and
> only the Douglas-Peucker algorithm is implemented.
> The appropriate level of topology is always maintained, centroids are
> always kept and never altered. Boundaries are not simplified if this
> would result in area deletion or changed centroid attachment. All the
> work is done with a temporary vector, and after simplifying the
> temporary vector, only alive lines are copied to the final output vector
> (no dead lines, smaller file size).  v.simplify can also delete
> duplicate points only, i.e. of two adjacent vertices in a line/boundary
> with identical coordinates one is removed. This should be done before
> simplification in v.simplify.
> As far as I understand the source code v.generalize discards all
> centroids and builds them anew at the end, area topology is not
> maintained during simplification.
> This is not meant to be competition for v.generalize, just some ideas on
> how to avoid dead lines in the output vector and how to preserve all
> areas/centroids, with their original category ID.
> 
> Markus
>  
> 
> Peter Löwe wrote:
>> Hi,
>>
>> I second the "unclean deletion" theory as I encountered this phenomenon before. Also, when generalizing areas which are attached to each other, while the overall "appearance" of the generalized borderlines is ok, more than 50 % of the areas cease to exist, which needs further investigation: Apart from "dead lines" we're dealing with "area mutilation" and "centroid abduction"!
>>
>> The (topological) truth is somewhere out there,
>> Peter
>>
>>
>> -------- Original-Nachricht --------
>>   
>>> Datum: Fri, 29 Aug 2008 19:17:46 +0300
>>> Von: Wolf Bergenheim <wolf+grass at bergenheim.net>
>>> An: dylan.beaudette at gmail.com
>>> CC: grass-user at lists.osgeo.org, hamish_b at yahoo.com, peter.loewe at gmx.de, Daniel Bundala <bundala at gmail.com>
>>> Betreff: Re: [GRASS-user] v.generalize / reanimation of dead lines ?
>>>     
>>   
>>> On 29.08.2008 18:54, Dylan Beaudette wrote:
>>>     
>>>> I have noticed this behavior as well when using v.generalize. I will try
>>>>       
>>> and 
>>>     
>>>> dig up the exact situation that caused it-- but I am pretty sure that
>>>>       
>>> the 
>>>     
>>>> initial linework was correct = unclean deletion.
>>>>
>>>>       
>>> That is very much possible. This module is still quite young and hasn't
>>> gone through a lot of live testing yet. Dylan, I'd be vey interested in
>>> this, if you can give me a simple case where it goes wrong.
>>>
>>> --Wolf
>>>
>>> Adding Daniel Bundala to the Cc list, as he wrote it last summer.
>>>
>>> -- 
>>>
>>> <:3 )---- Wolf Bergenheim ----( 8:>
>>>     
>>   
> _______________________________________________
> grass-user mailing list
> grass-user at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-user
> 
> 


-- 
Benjamin Ducke
Senior Applications Support and Development Officer

Oxford Archaeology Ltd
Janus House
Osney Mead
OX2 0ES
Oxford, U.K.

Tel: +44 (0)1865 263 800 (switchboard)
Tel: +44 (0)1865 980 758 (direct)
Fax :+44 (0)1865 793 496
benjamin.ducke at oxfordarch.co.uk




------
Files attached to this email may be in ISO 26300 format (OASIS Open Document Format). If you have difficulty opening them, please visit http://iso26300.info for more information.