Evaluation of rectification error.

[Andrea Giacomelli]

ons???Apr??3???9?59am?@bnr.ca>
content-type: text/plain; charset=us-ascii
organization: University of Illinois at Urbana
mime-version: 1.0
reply-to: grassu at isis.cecer.army.mil
newsgroups: info.grass.user
originator: daemon at ux1.cso.uiuc.edu



Hi,

When I use i.points, I am given for each control point a corresponding
value for the rectification error.

The concern I have is that if I perform a rectification, that figure
applies only to that specific point, while in areas which are far from
GCPs (not always it's possible to obtain a well distributed set of them...),
I expect the error to be larger.

I would like to quantify this someway, but I'm not a cartographer, so
I have to learn bits & pieces on the way: would it make sense to
interpolate the rms errors for the single points over the rectified area ?
If so, what type of algorithm would be safer (s.surf.idw, s.surf.tp, or else ?)

Thank you,

-- 
                        Andrea Giacomelli
                        andreag at idra6.iar.polimi.it
                        http://idra6.iar.polimi.it:8080/~andreag
                        tel: +39-2-23996233 fax: +39-2-23996207