«

(brian) wrote:

> Also, the world's most expensive panoramic camera doesn't appear to
> have the ability to keep the entrance pupil of the lens fixed while
> making pitch and yaw rotations. Groan.

I thought all panorama software estimated the projections of the
original images and then re-projected them onto the larger canvas.
Wouldn't it have to? There is no way Random J. Photographer with a
handheld camera is going to be able to maintain an accurate entrance
pupil while snapping off the source images.

The JPL images look like they were assembled by hand, in a hurry, with
material they had available (some of it of questionable quality).
Re-shoots aren't easy when the pipe is a ~hundred million kilometres
long and tens(?) of kilobits/second wide.

(brian) writes:

> Mark Herring <> wrote in message news:<>. ..
> > On 5 Jan 2004 12:05:28 -0800, (brian) wrote:
> >
> > >When will NASA get wind of Panorama Tools so they can do things
> > >right?? http://marsrovers.jpl.nasa.gov/galle...a/PIA04991.jpg
> > >
> > >Also, the world's most expensive panoramic camera doesn't appear to
> > >have the ability to keep the entrance pupil of the lens fixed while
> > >making pitch and yaw rotations. Groan.

Perhaps because it's a pair of cameras for stereo. It's hard to keep both
camera viewpoints fixed and still make a full-circle panorama .

<snip>

--
-Stephen H. Westin
Any information or opinions in this message are mine: they do not
represent the position of Cornell University or any of its sponsors.

wrote in message news:< om>...
> (brian) wrote:
>
> > Also, the world's most expensive panoramic camera doesn't appear to
> > have the ability to keep the entrance pupil of the lens fixed while
> > making pitch and yaw rotations. Groan.
>
> I thought all panorama software estimated the projections of the
> original images and then re-projected them onto the larger canvas.
> Wouldn't it have to? There is no way Random J. Photographer with a
> handheld camera is going to be able to maintain an accurate entrance
> pupil while snapping off the source images.
>
> The JPL images look like they were assembled by hand, in a hurry, with
> material they had available (some of it of questionable quality).
> Re-shoots aren't easy when the pipe is a ~hundred million kilometres
> long and tens(?) of kilobits/second wide.


Go ahead and download Maestro at http://mars.telascience.org/

You'll be able to manipulate the raw images provided by the rovers.
It'll also create anaglyphs for you to take a gander at...using those
50s-like 3d glasses.

wrote in message news:< om>...
> (brian) wrote:
>
> > Also, the world's most expensive panoramic camera doesn't appear to
> > have the ability to keep the entrance pupil of the lens fixed while
> > making pitch and yaw rotations. Groan.
>
> I thought all panorama software estimated the projections of the
> original images and then re-projected them onto the larger canvas.
> Wouldn't it have to? There is no way Random J. Photographer with a
> handheld camera is going to be able to maintain an accurate entrance
> pupil while snapping off the source images.
>
> The JPL images look like they were assembled by hand, in a hurry, with
> material they had available (some of it of questionable quality).
> Re-shoots aren't easy when the pipe is a ~hundred million kilometres
> long and tens(?) of kilobits/second wide.


You've got the basic algorithm correct, and with proper technique the
stitching errors can be brought to less than 1 pixel even for
extremely large mosaics. Of course, proper technique generally
requires a tripod and pano head unless everything is effectively at
infinity. Hopefully you are right about the preliminary nature of the
images, and future ones will be taken and processed correctly.

Brian
www.caldwellphotographic.com

wrote:

> (brian) wrote:
>
>
>>Also, the world's most expensive panoramic camera doesn't appear to
>>have the ability to keep the entrance pupil of the lens fixed while
>>making pitch and yaw rotations. Groan.
>
>
> I thought all panorama software estimated the projections of the
> original images and then re-projected them onto the larger canvas.
> Wouldn't it have to? There is no way Random J. Photographer with a
> handheld camera is going to be able to maintain an accurate entrance
> pupil while snapping off the source images.
>
> The JPL images look like they were assembled by hand, in a hurry, with
> material they had available (some of it of questionable quality).
> Re-shoots aren't easy when the pipe is a ~hundred million kilometres
> long and tens(?) of kilobits/second wide.

They look like manual pasteups to me. I suspect that they are in quite
a rush to get them out. I am SURE they have software that can do a
better job of stiching that what I have seen so far...

kpfeif wrote:

> wrote in message news:< om>...
>
>> (brian) wrote:
>>
>>
>>>Also, the world's most expensive panoramic camera doesn't appear to
>>>have the ability to keep the entrance pupil of the lens fixed while
>>>making pitch and yaw rotations. Groan.
>>
>>I thought all panorama software estimated the projections of the
>>original images and then re-projected them onto the larger canvas.
>>Wouldn't it have to? There is no way Random J. Photographer with a
>>handheld camera is going to be able to maintain an accurate entrance
>>pupil while snapping off the source images.
>>
>>The JPL images look like they were assembled by hand, in a hurry, with
>>material they had available (some of it of questionable quality).
>>Re-shoots aren't easy when the pipe is a ~hundred million kilometres
>>long and tens(?) of kilobits/second wide.
>
>
>
> Go ahead and download Maestro at http://mars.telascience.org/
>
> You'll be able to manipulate the raw images provided by the rovers.
> It'll also create anaglyphs for you to take a gander at...using those
> 50s-like 3d glasses.

I did, but where do I go for the raw images?

In article < >,
(kpfeif) wrote:

> Go ahead and download Maestro at http://mars.telascience.org/

> You'll be able to manipulate the raw images provided by the rovers.
> It'll also create anaglyphs for you to take a gander at...using those
> 50s-like 3d glasses.


I've tried Maestro in "tutorial" mode.
On my computer the rover has large blank areas under it and there is
terrain on top of the solar panels.
How does it look on yours?
I'm wondering if it is their problem or mine.
--
Charlie Dilks
Newark, DE USA