[GRASS-user] ridge extraction from DEM
Michael Barton
michael.barton at asu.edu
Thu Jun 14 09:00:27 PDT 2012
Given your objectives, I think r.param.scale or r.prominence would be your best choices.
Michael
On Jun 14, 2012, at 2:29 AM, Thomas Lee wrote:
> Hi Michael
>
> My approach of voting is trying to solve the subjective human perceptaion of landforms per culture.
>
> Voting here means sample population test at defined culture such as
> Culture 1, Sample size n = 300, Mean = 5, S.D. = 2, Sampling Time = T1
> Culture 2, Sample size n = 500, Mean = 4, S.D. = 3, Sampling Time = T2
> Culture 3, Sample size n = 1000, Mean = 6, S.D. = 1, Sampling Time = T3
>
> They are dynamic and can be stored in a database.
> They are temporal information (voting information), subject to change per year (or per month depended on the attribute sensitivity to the time), per culture (spatial location).
> It is better put in a web database server and update globally and remotely by mobile device such as android phone and iphone.
>
> All come together as a knowledge base (I am used to do it in my M.Sc. thesis, about 20 years ago, on spatial decision support system for location-allocation analysis using concept of expert system, knowledge base, model base, data base.. etc)
>
> Thanks for introducing "Prominence".
> It is same as the Feng Shui "Mountain Ridge" system logic (or Form School).
> The masters have been studying these for, it is beleived that, thousands of years.
> Their concept is to locate the best site by studying the lineage of prominence.
> For country capital, they focus in top range or high level hierarchies lineage of prominence
> For smaller captial such as provincial capital, they use middle range lineage
> For small city or village, they use lower range lineage.
>
> The spot they want to locate is likely a flat area at the end of a gentle slope ridge line with big openning in front and surrounded by embracing ridges.
> With this pattern, stream lines will be converged in front of the site and having high quality of soil content suitable for substaintable living.
>
> I consider this logic can be proved by analyzing
> - the prominence network
> - length of lineage
> - height of parent and subpeak
> - total lengeth of river/stream network
> - accumulated flow of water (accumulate rich content in water)
> - accumulated length of ridge line (in Feng Shui, they interested in the Energy (Qi) bring along through the mountain ridge line, longer the line, high energy, better benefit to the site)
> - formation of a complete mountain hierachies of line parent and subpeak (stored in a mountain hierachy database)
> - entropy of peak (distribution of peak)
> - ridge pattern recognition (Feng Shui has its mountain dragon (ridge) pattern system such as Lotus Pattern)
>
> The MH database
> - Table of Mountain, point layer
> - Table of lineage, line layer (route structure like road network with junction as peak or subpeak)
> - Table of stream node, point layer
> - Table of stream/river, line layer (route structure)
> - Table of event (like road side event such as accident, but here is Feng Shui site, village, capital, city)
>
> Thomas
>
>
> ----- Original Message -----
> From: Michael Barton
> To: Thomas Lee
> Sent: Thursday, 14 June, 2012 1:43 AM
> Subject: Re: 回覆： Re: [GRASS-user] 回覆： Re: ridge extraction from DEM
>
> Hi Thomas,
>
> You can indeed come up with quantitative definitions. My point is that these will have a variable match to subjective human perception of landforms. The ranking you list below can be applied to any topographic prominence, including landforms called hills, bluffs, ridges, etc. The question of ridges (as long, relatively narrow landforms) is slightly different from one of topographic prominence. There is in fact a module (I think it is called r.prominence) in addons that will ID topographic prominence by various characteristics that you can set. Maybe it would be of use to you.
>
> Michael
>
>
> On Jun 13, 2012, at 10:20 AM, Thomas Lee wrote:
>
>> Michael
>>
>> The qualitative definition of ridge can be quantitatively modeled as
>>
>> size in metre; degree of size of ridge
>> 1; 1
>> 5; 2
>> 10; 3
>> 50; 4
>> 100; 5
>> .....
>> 1000; 10
>>
>> Normally a flat area bigger than 1000m is a platform where a small village can be accomodated.
>> The above can be modelled by fuzzy membership function which can be implemented in GRASS
>>
>> By having questionaire in different cultures at different locations globally, the result could be like voting in expert system voting model for spatial decision support system.
>>
>> I am interested in the correlation of the spatial cluster of human settlement to the pattern of ridges. Not a ridge but spatial pattern of a group of ridges.
>> It is going to test the Chinese Feng Shui concept. Feng Shui is the ancient technique in site selection. To find a site good for living over say hundreds or thousands of years for next generations based on geomorphically terrain pattern or Mountain Form. There are systems in Feng Shui where the Master study the ridge pattern (and stream pattern) to find the site with good soil content without using modern borehole or lab test technique.
>>
>> Thomas Lee
>>
>>
>>
>> ----- Original Message -----
>> From: Michael Barton
>> To: Thomaswplee
>> Cc: markus.metz.giswork at googlemail.com ; diregola at gmail.com ; grass-user at lists.osgeo.org
>> Sent: Wednesday, 13 June, 2012 1:39 AM
>> Subject: Re: 回覆： Re: [GRASS-user] 回覆： Re: ridge extraction from DEM
>>
>> Thomas,
>>
>> Again, you'd have to look to see if this can give you the results you need.
>>
>> The word "ridge" is a qualitative term that refers to a general landform category. For your purposes, how high does a landform need to be to qualify as a ridge: 1m, 10m, 100m, 1000m? Similarly, how long does it need to be and what is the acceptable range of profile slope along a ridge-top?
>>
>> These feature extraction methods deal in numbers, so they will probably never exactly match the subjective perception of a landform as a ridge. This is especially true if you are interested in how people (in the past and in the present) perceive a landform and its suitability for settlement. But you should be able to get reasonably close to some kind of consistent perception of what is a ridge, and the GIS methods have the advantage of being explicit, transparent, and repeatable--essential for science of course. But there is probably no "best" method for ID of ridges.
>>
>> I haven't tried the convergence method, but will do so to see how it turns out.
>>
>> Michael
>>
>> On Jun 12, 2012, at 5:41 AM, Thomaswplee wrote:
>>
>>> What is the difference on convergence and ridge?
>>> Both local convex quadratically double differentiated
>>>
>>> I am interesting in application of geomorphology vs social science such as terrain analysis related to human settlement
>>>
>>> Thomas
>>>
>>>
>>>
>>> -------- Original message --------
>>> Subject: Re: [GRASS-user] 回覆： Re: ridge extraction from DEM
>>> From: Markus Metz <markus.metz.giswork at googlemail.com>
>>> To: Margherita Di Leo <diregola at gmail.com>
>>> CC: Michael Barton <michael.barton at asu.edu>,grass-user grass-user <grass-user at lists.osgeo.org>,Thomaswplee <thomaswplee at gmail.com>
>>>
>>>
>>> On Tue, Jun 12, 2012 at 11:28 AM, Margherita Di Leo <diregola at gmail.com> wrote:
>>> > Hi,
>>> >
>>> > there is also r.convergence add-on:
>>> > http://grass.osgeo.org/wiki/GRASS_AddOns#r.convergence
>>>
>>> BTW, the topographic convergence index is also available in
>>> r.watershed in GRASS 7.
>>>
>>> Markus M
>>>
>>> >
>>> > Regards,
>>> > madi
>>> >
>>> > On Tue, Jun 12, 2012 at 4:24 AM, Michael Barton <michael.barton at asu.edu>
>>> > wrote:
>>> >>
>>> >> Invert the DEM by multiplying all values by -1 and adding the maximum
>>> >> original height value. This makes the maximum height 0 and everything less
>>> >> than that increasingly greater than 0
>>> >>
>>> >> Run r.watershed on the inverted DEM choosing stream segments as output.
>>> >>
>>> >> The stream segments from the inverted DEM are your ridgelines. You can
>>> >> keep them in raster or thin them (r.thin) and convert them to vector
>>> >> (r.to.vect).
>>> >>
>>> >> I've copied the GRASS user list again because this is a general question
>>> >> that others might be interested in.
>>> >>
>>> >> Michael
>>> >>
>>> >>
>>> >>
>>> >> On Jun 11, 2012, at 7:29 PM, Thomaswplee wrote:
>>> >>
>>> >> How to make ridge vector line with reversed accumlated flow as in
>>> >> watershed?
>>> >>
>>> >> Still reverse DEM plus vectorization of ridge raster?
>>> >>
>>> >> Thomas
>>> >>
>>> >>
>>> >>
>>> >> -------- Original message --------
>>> >> Subject: Re: ridge extraction from DEM
>>> >> From: Michael Barton <michael.barton at asu.edu>
>>> >> To: thomaslee at starvision.com.hk
>>> >> CC: grass-user grass-user <grass-user at lists.osgeo.org>
>>> >>
>>> >>
>>> >> Thomas,
>>> >>
>>> >> I don't remember posting anything about r.ppa. I did a quick search and
>>> >> found a post by Māris Nartišs
>>> >> (http://lists.osgeo.org/pipermail/grass-user/2005-October/030883.html).
>>> >>
>>> >> r.param.scale will extract ridges. You will need to adjust the optional
>>> >> parameters (especially processing window size) to get the ridges you want.
>>> >>
>>> >> An alternative method is to invert the DEM and run r.watershed on it,
>>> >> extracting the 'streams'. The 'streams' of the inverted DEM will follow
>>> >> ridges.
>>> >>
>>> >> Michael
>>> >>
>>> >>
>>> >>
>>> >> On Jun 11, 2012, at 9:26 AM, <thomaslee at starvision.com.hk>
>>> >> wrote:
>>> >>
>>> >> > Dear Michael,
>>> >> >
>>> >> > I am interested in ridge extraction but not able to find the r.ppa
>>> >> > mentioned by
>>> >> >
>>> >> >
>>> >> > Chang, Y.C., Frigeri, A., 2002. Implementing the automatic
>>> >> > extraction of ridge and valley axes using the PPA algorithm in
>>> >> > Grass GIS. In: Open Source Free Software GIS GRASS
>>> >> > Users Conference, 2002.
>>> >> >
>>> >> >
>>> >> > Implementing the automatic extraction of ridge and
>>> >> > valley axes using the PPA algorithm in Grass GIS
>>> >> >
>>> >> >
>>> >> > Do you have any solution since your message in 2007 about ridge
>>> >> > extraction.
>>> >>
>>> >> _____________________
>>> >> C. Michael Barton
>>> >> Visiting Scientist, Integrated Science Program
>>> >> National Center for Atmospheric Research &
>>> >> University Corporation for Atmospheric Research
>>> >> 303-497-2889 (voice)
>>> >>
>>> >> Director, Center for Social Dynamics & Complexity
>>> >> Professor of Anthropology, School of Human Evolution & Social Change
>>> >> Arizona State University
>>> >> www: http://www.public.asu.edu/~cmbarton, http://csdc.asu.edu
>>> >>
>>> >>
>>> >> _____________________
>>> >> C. Michael Barton
>>> >> Visiting Scientist, Integrated Science Program
>>> >> National Center for Atmospheric Research &
>>> >> University Consortium for Atmospheric Research
>>> >> 303-497-2889 (voice)
>>> >>
>>> >> Director, Center for Social Dynamics & Complexity
>>> >> Professor of Anthropology, School of Human Evolution & Social Change
>>> >> Arizona State University
>>> >> www: http://www.public.asu.edu/~cmbarton, http://csdc.asu.edu
>>> >>
>>> >>
>>> >>
>>> >>
>>> >>
>>> >>
>>> >> _______________________________________________
>>> >> grass-user mailing list
>>> >> grass-user at lists.osgeo.org
>>> >> http://lists.osgeo.org/mailman/listinfo/grass-user
>>> >>
>>> >
>>> >
>>> >
>>> > --
>>> > Dr. Margherita Di Leo
>>> >
>>> > _______________________________________________
>>> > grass-user mailing list
>>> > grass-user at lists.osgeo.org
>>> > http://lists.osgeo.org/mailman/listinfo/grass-user
>>> >
>>
>> _____________________
>> C. Michael Barton
>> Visiting Scientist, Integrated Science Program
>> National Center for Atmospheric Research &
>> University Corporation for Atmospheric Research
>> 303-497-2889 (voice)
>>
>> Director, Center for Social Dynamics & Complexity
>> Professor of Anthropology, School of Human Evolution & Social Change
>> Arizona State University
>> www: http://www.public.asu.edu/~cmbarton, http://csdc.asu.edu
>>
>>
>
> _____________________
> C. Michael Barton
> Visiting Scientist, Integrated Science Program
> National Center for Atmospheric Research &
> University Consortium for Atmospheric Research
> 303-497-2889 (voice)
>
> Director, Center for Social Dynamics & Complexity
> Professor of Anthropology, School of Human Evolution & Social Change
> Arizona State University
> www: http://www.public.asu.edu/~cmbarton, http://csdc.asu.edu
>
>
>
>
>
>
_____________________
C. Michael Barton
Visiting Scientist, Integrated Science Program
National Center for Atmospheric Research &
University Corporation for Atmospheric Research
303-497-2889 (voice)
Director, Center for Social Dynamics & Complexity
Professor of Anthropology, School of Human Evolution & Social Change
Arizona State University
www: http://www.public.asu.edu/~cmbarton, http://csdc.asu.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-user/attachments/20120614/2c2dbd1c/attachment-0001.html>
More information about the grass-user
mailing list