Wild imaging ideas #1

Planets rotate while you are trying to image them.
If you image a planet, Jupiter for instance, for a periods of many
hours, maybe even many days, the imaqges would all be of a slightly
different image. The surface details would be blured by the rotation.
Why not correct for this? Take the individual round images and
project them onto a sphere, index for rotation and then stack the
result.
One could even observe for days on end and the surface model would get
better and better?


Is this idea Daft?


Paul

"Paul Breed" wrote in message
...
Planets rotate while you are trying to image them.
If you image a planet, Jupiter for instance, for a periods of many
hours, maybe even many days, the imaqges would all be of a slightly
different image. The surface details would be blured by the rotation.
Why not correct for this? Take the individual round images and
project them onto a sphere, index for rotation and then stack the
result.
One could even observe for days on end and the surface model would get
better and better?


Is this idea Daft?


Paul


A planet, such as Jupiter, is very bright and requires only a few seconds to
capture an image. I can't see any purpose or advantage in extending the
exposure into hours or days. Besides, the image would be over exposed in a
matter of seconds.

Al

"Paul Breed" wrote in message
...
Planets rotate while you are trying to image them.
If you image a planet, Jupiter for instance, for a periods of many
hours, maybe even many days, the imaqges would all be of a slightly
different image. The surface details would be blured by the rotation.
Why not correct for this? Take the individual round images and
project them onto a sphere, index for rotation and then stack the
result.
One could even observe for days on end and the surface model would get
better and better?

Actually I wish somebody would write a computer program to take my Mars
images (with known central latitude and longitude) and "unwrap" them into a
standard map projection. I could then stitch them together to make a map of
Mars.

This is not a silly task. Because of sand blowing around in the wind, the
albedo (brightness) features of Mars change from year to year, and it's a
good idea to make a new albedo map every time Mars comes to opposition.

Take and stack thousands of images to improve the resolution.
Use the rotation model to eliminate the rotation errors in the
"Stacking"

On Tue, 16 Sep 2003 03:56:05 GMT, "Al"
wrote:


"Paul Breed" wrote in message
.. .
Planets rotate while you are trying to image them.
If you image a planet, Jupiter for instance, for a periods of many
hours, maybe even many days, the imaqges would all be of a slightly
different image. The surface details would be blured by the rotation.
Why not correct for this? Take the individual round images and
project them onto a sphere, index for rotation and then stack the
result.
One could even observe for days on end and the surface model would get
better and better?


Is this idea Daft?


Paul


A planet, such as Jupiter, is very bright and requires only a few seconds to
capture an image. I can't see any purpose or advantage in extending the
exposure into hours or days. Besides, the image would be over exposed in a
matter of seconds.

Al

Is this idea Daft?

YES.

Paul Breed wrote:
Planets rotate while you are trying to image them.
If you image a planet, Jupiter for instance, for a periods of many
hours, maybe even many days, the imaqges would all be of a slightly
different image. The surface details would be blured by the rotation.
Why not correct for this? Take the individual round images and
project them onto a sphere, index for rotation and then stack the
result.
One could even observe for days on end and the surface model would get
better and better?


Is this idea Daft?


Paul
Paul,

Think about doing this for the earth, say from earth orbit... like a
weather satellite. Your plan would work except the weather patterns
change. Apart from that, taking images over a single night would
probably work quite well. Imaging over a few nights might give too much
movement to the weather patterns themselves.

Eric.

Paul Breed wrote in message . ..
Planets rotate while you are trying to image them.
If you image a planet, Jupiter for instance, for a periods of many
hours, maybe even many days, the imaqges would all be of a slightly
different image. The surface details would be blured by the rotation.
Why not correct for this? Take the individual round images and
project them onto a sphere, index for rotation and then stack the
result.
One could even observe for days on end and the surface model would get
better and better?


Is this idea Daft?

The basic idea is brilliant, and has (though in a different guise) been in
use for some time. What you have described is basically the very foundation
of Synthetic Aperture Radar (SAR), which has been available since the 1970ies:
The radar platform takes numerous "snapshots" of the target (e.g. a planet)
surface from slightly different angles, correct for the differences in
offsets and use all this information to build an increasingly better image.

Now, while the basics may be OK, the practical implementation may be
problematic. Extending the synthetic aperture technique to the sonar case
is, from a mathemathical point of view, trivial: Put in the speed of sound
instead of the speed of light, then change some carrier frequencies and pulse
bandwidths. Unfortunately, operating a sonar platform under water is not
quite as easy as operating a radar platform (aircraft or satellite), since
the underwater currents and waves tend to disturb the sonar platform quite
severly. Another problem is that the speed of sound in water varies
dynamically on the order of 2-5 %, which means it is very difficult to
process the recorded echoes to extract the desired level of detail. In fact,
the earliest account I have seen on the principles of Synthetic Aperture
Sonar dates from 1976, but practical implementations (concept demonstrations)
have only appeared in the last two or three years.

So you need to check out if the object to be imaged as well as other
circumstances are favourable for a given imaging thechnique. As for the
Jupiter image, there will be some sort of "longest practical observation
window", i.e. a time scale where the features of the planet, like patterns
in storms and turbulent areas, change significantly. In the Jupiter case,
I suspect we will be talking hours at most. Stacking images gathered on
longer time frames than this window will probably only produce blurred
images anyway. So the question is whether one just as well should take one
long exposure during this time frame.

Having said that, your proposed method could very well be used for rigid,
cloudless planets. Perhaps Mars in periods with good Mars weather.

Rune

Michael A. Covington va escriu


"Paul Breed" wrote in message
...
Planets rotate while you are trying to image them.
If you image a planet, Jupiter for instance, for a periods of many
hours, maybe even many days, the imaqges would all be of a slightly
different image. The surface details would be blured by the rotation.
Why not correct for this? Take the individual round images and
project them onto a sphere, index for rotation and then stack the
result.
One could even observe for days on end and the surface model would get
better and better?

Actually I wish somebody would write a computer program to take my Mars
images (with known central latitude and longitude) and "unwrap" them into
a
standard map projection. I could then stitch them together to make a map
of Mars.

This is not a silly task. Because of sand blowing around in the wind, the
albedo (brightness) features of Mars change from year to year, and it's a
good idea to make a new albedo map every time Mars comes to opposition.

Check what Thierry Legault did for the images of the Shoemaker-Levy 9
collision with Jupiter.

http://perso.club-internet.fr/legault/

Regards,

--
/************************************************** **/
/* */
/* Víctor Zabalza */
/* */
/* vzabalza @ teleline.es vzabalto7 @ fis.ub.edu */
/* */
/* GNU/Linux Debian Sarge - Kernel v2.6.0-test5 */
/* */
/************************************************** **/

NO a la guerra!! NO en el nostre nom, NO amb el nostre silenci!!

Fortune del dia:

No se está realmente con los pobres sino luchando contra la pobreza.
-- Paul Ricoeur.

Eric wrote:
Paul Breed wrote:
Planets rotate while you are trying to image them.
If you image a planet, Jupiter for instance, for a periods of many
hours, maybe even many days, the imaqges would all be of a slightly
different image. The surface details would be blured by the rotation.
Why not correct for this? Take the individual round images and
project them onto a sphere, index for rotation and then stack the
result.
One could even observe for days on end and the surface model would get
better and better?


Is this idea Daft?


Paul
Paul,

Think about doing this for the earth, say from earth orbit... like a
weather satellite. Your plan would work except the weather patterns
change. Apart from that, taking images over a single night would
probably work quite well. Imaging over a few nights might give too much
movement to the weather patterns themselves.

In certain cases the notion works - it's close to the way Marc Buie
mapped Pluto with HST data. The wrinkle in that cases was that the
resolution wasn't much better than the disk size, so each new image
improved an iterative model which was projected and blurred to
match the viewing geometry in each case.

Bill Kell