[GRASSLIST:5184] RE: Regarding interpolation

[Antonio Uriarte]

Hello.

----- Original Message -----
From: Abhay Kulkarni <abhayak_anr at sancharnet.in>
To: <GRASSLIST at baylor.edu>
Sent: Sunday, December 08, 2002 5:43 AM
Subject: [GRASSLIST:5182] Regarding interpolation


> Hi all !!
>
> I am a new grassuser,totally new to GIS and working on grass5.0.0pre2_i686
on
> my RH Linux 7.1. Basically I am a computer profession but interested a lot
in
> Geography.
>
> I digitized a contour map of the scale 1 cm = 30 mtrs. (1:3000) in x-y
> projection. (image size 3500 x 2550 pixels , 300dpi). I wanted to
visualize
> it in 3D. The modules r.thin,r.line,r.contour could not help as
(probably )
> there was a lot of text and other markings on the scanned image. So I
> digitized the map manually using v.digit and attributed the lines.
>
> When I ran the module v.surf.rst, it gave the message " taking too long to
> find points for interpolation - please change the region to area where
your
> points are. Continuing calculations... ". Before I understand its meaning,
> the system hanged. This happened twice.

I suposse that you have defined a region too big and there is a wide empty
frame around the area you digitized. You have to redefine your region
boundaries (g.region) and adjust them to the area your lines are.

>
> Then I ran v.support and then v.to.rast. Then I used r.surf.idw2, created
the
> elevation map and viewed it in nviz. I got a wonderful 3D view. That
matched
> the contour lines but shape was not smooth (like elevation.dem in the
> spearfish). It was like polygons of sharp edges stacked on each other.

r.surf.idw2 is not a good method for interpolating contours. Try with
rasterized contours (r.surf.contour).
>
> Will anybody please guide me where I was wrong ? It took some 2 and a half
> hours for r.surf.idw2 on my 128 MB (plus 256 Swap) PC for 1256 rows. Is
this
> a normal time ?

Yes, it is a normal time. Interpolation uses to take long.
>
> Thanking you.
>
> Abhay Kulkarni,
> India.
>