[GRASSLIST:5048] difficulties with i.ortho.photo

Hello,

I am using i.ortho.photo to ortho-rectify aerial images of Mayotte (Comoros Archipelago).
The associated man pages are not very clear about a few points.

When i register a camera, what is this "point of symmetry" that i have to locate on the photo ? What is the origin and direction of X and Y on the photo ?

When i must enter fiducial and reseau marks, where is the origin ? is it the same as the point of symmetry that i talked about above or is the point of symmetry the new origin ?

Well, if someone has experience in using i.ortho.photo and explain me a bit, i'd be very grateful.

Thanks in advance,

Guillaume Criloux.

When i register a camera, what is this "point of symmetry" that i have to locate on the photo ?

It is something that is determined in the camera calibration. The photogrammetrist should be able to give you a calibration report for the camera, but if you don't know, assume 0,0. It's the center of the image.

What is the origin and direction of X and Y on the photo ?

The origin is the center of the image (or the point of symetry) and X and Y are left-right and up-down.

My fiducial marks look something like this:
   X Y
106 -106 lower right corner
-106 -106 lower left corner
-106 106 upper left corner
106 106 upper right corner
  0 112 bottom middle
-112 0 left middle
   0 112 top middle
112 0 right middle

Rich

Richard W. Greenwood, PLS
(307) 733-0203
Rich@GreenwoodMap.com
www.GreenwoodMap.com

Hi Guillaume,

Here is a quick diagram to explain what's required for proceeding to
the interior orientation of your air-photos (that's the photogrammetric
jargon for what you want to do).

The marks on the side of the photos are called fiducial marks (FM).
If you want a good interior orientation, you'd better know the order in
which they were calibrated. Look at the direction of the axes X and
Y. In the diagram, the Y axis has got positive values as it goes
down. This may not be the case on your images and depends on
the lab where the camera calibration was done. My diagram is
proper to the camera Leica LMK9 with serial number 265-002A
belonging to the Greek Army. You should see the serial number of
the camera in the instrument strip (black band where you can see
the altimeter, watch and photo counter of the camera). As a crude
indication, the instrument strip bears the so-called principal distance
(also known as focal length of the camera). The actual calibrated
principal distance will be slightly different than this one. You should
use the calibrated value.

Note that the order in which you enter the fiducials will determine
virtually the direction in which the plane was flying. Follow the order
of the calibration certificate. There should be a small hand-
annotated diagram with the fiducials numbering scheme.

The centre of symmetry is the point marked PP. Ideally it should fall
bang on the intersection of the radii at the centre of the picture, but
they never do. The excentricity as it is called is usually of the order
of a few micrometers. The origin of the axes is at the intersection of
the radii traced from the fiducial marks.

Your best bet to solve you problem is to get hold of the calibration
certificate of the camera that was used to shoot the photos. The
photo retailer should provide it to you for free. Also, unless it is for
illustration purposes, the photographs must scanned on a proper
photogrammetric scanner otherwise you'll end up with unexplained
distortions in your orthophotos.

Does this help?
Best,

Thomas

Hello,

I am using i.ortho.photo to ortho-rectify aerial images of Mayotte
(Comoros Archipelago). The associated man pages are not very clear
about a few points.

When i register a camera, what is this "point of symmetry" that i have
to locate on the photo ? What is the origin and direction of X and Y
on the photo ?

When i must enter fiducial and reseau marks, where is the origin ? is
it the same as the point of symmetry that i talked about above or is
the point of symmetry the new origin ?

Well, if someone has experience in using i.ortho.photo and explain me
a bit, i'd be very grateful.

Thanks in advance,

Guillaume Criloux.

Thomas Dewez
PhD Student
Dept Geography & Earth Sciences
Brunel University (West London)
Uxbridge UB8 3PH, Mddx
United Kingdom

Phone: +44-(0)1895-203215
Fax: +44-(0)1895-203217

e-mail: thomas.dewez@brunel.ac.uk

The following section of this message contains a file attachment
prepared for transmission using the Internet MIME message format.
If you are using Pegasus Mail, or any another MIME-compliant system,
you should be able to save it or view it from within your mailer.
If you cannot, please ask your system administrator for assistance.

   ---- File information -----------
     File: Airphoto-sketch.gif
     Date: 5 Aug 2002, 10:25
     Size: 13973 bytes.
     Type: GIF-image

(attachments)

Airphoto-sketch.gif

On Sun, Nov 24, 2002 at 01:06:30PM +0300, Guillaume Criloux wrote:

Hello,

I am using i.ortho.photo to ortho-rectify aerial images of Mayotte (Comoros Archipelago).
The associated man pages are not very clear about a few points.

When i register a camera, what is this "point of symmetry" that i have to locate on the photo ? What is the origin and direction of X and Y on the photo ?

When i must enter fiducial and reseau marks, where is the origin ? is it the same as the point of symmetry that i talked about above or is the point of symmetry the new origin ?

Well, if someone has experience in using i.ortho.photo and explain me a bit, i'd be very grateful.

Please have a look at
http://grass.itc.it/gdp/html_grass5/html/photo.camera.html

I have added some explanations. Please suggest if this is not sufficient.

Best regards
Markus Neteler

Hello all

does anyone know if there is the possibility to orthorectify high resolution
saltellite images, i.e. QuickBird and Ikonos.

Thank to all

Bye

On Wed, Nov 27, 2002 at 04:30:58PM +0100, Ivan Marchesini wrote:

Hello all

does anyone know if there is the possibility to orthorectify high resolution
saltellite images, i.e. QuickBird and Ikonos.

A LANDSAT orthorectification is implemented in
i.points3
i.rectify3
src/libes/image3/

But this is untested (testers wanted) and probably not fully
implemented (testers wanted).

The implementation *may* work also for other platforms than LANDSAT (no
idea).

Cheers
Markus Neteler