i.rectify

[Jean-Marc Chofflet]

I asked a question regarding the use of i.rectify(2) and r.patch

> My second problem is less obvious : I've scanned a map in 15 images.
> Before digitizing it (with another software), I've i.rectified each of
> them
> (i.rectify2, 3rd order, at least 30 GCP's per image). When I try to
> put the 15 rectified images together (after an r.out.tga) in any
> software 
> (I'm now using CorelXARA under Windows) there are some distorsions, but
> if
> I first r.patch those and then r.out.tga the result, it fits perfectly!
> I thought the images produced by i.rectify2 were... rectified!!!
> You could say : "never mind, use r.patch!". Yes, except that it would
> produce a file my poor Pentium 133, 64 MB of RAM and 4GB of HDD wouldn't
> dare
> to just look at it!

Malcolm answered me :

>Could you give a little more information about the images that you are 
>scanning? Unless they are non-ortho-rectified aerial photos, chances are 
>that you really only need a first order rectification, simply to 
>geo-reference them. I'm really unclear about whether you were r.patch-ing 
>them before or after rectification.


The images I'm using are the scanned images of a paper map I want to
digitize. Since the sheet is larger then my scanner, I've produced 15
images.
I've used i.rectify2 since the resulting images are not perfectly
accurate
(my map is an old one with many folds...).

My first step was to digitize each of the images in CorelXARA (vectorial
software)
and try to fit the pieces together : some distorsions were too important
so I've used
i.rectify2 under GRASS, then produced 15 rectified images. I've tried
again under CorelXARA :
again some distorsions (not the same ones). So I r.patched my rectified
images to produce
one image : it works (almost perfectly)! My question is : does r.patch
do more then just 
connecting the pieces together or did I misused CorelXARA when I tried
to join the rectified
images "by hand"?

 Thanks again


***************************************************
*                                                 *
*     JM Chofflet                                 *
*                                                 *
*     ERA 41 du CNRS       7, rue Gaston Couté    *
*     3, rue Michelet      F75018 Paris           *
*     F75006 Paris                                *
*                                                 *
*     tel off. : +33 (0)1 53 73 71 08             *
*     tel home : +33 (0)1 46 06 88 59             *
*     fax      : +33 (0)1 44 07 01 79             *
*     Mailto:chofflet at asterix.univ-paris1.fr      *
*                                                 *
***************************************************