Mailinglist Archives:
Infrared
Panorama
Photo-3D
Tech-3D
Sell-3D
MF3D
|
|
Notice |
This mailinglist archive is frozen since May 2001, i.e. it will stay online but will not be updated.
|
|
[photo-3d] Stereo Panorama Geometry
- From: John Toeppen <toeppen@xxxxxxxx>
- Subject: [photo-3d] Stereo Panorama Geometry
- Date: Wed, 13 Sep 2000 22:35:13 -0700
Ron is right in saying that a panning camera should piviot around a
point just behind the lens in using a 50 mm fl SLR lens. This will
produce the best "tape the prints" mono panorama. Use of a slide bar
violates this condition. Unfortuantely, it also produces some
objectionable perspective offsets unless many, many photos are taken.
A single camera may be mounted on a tripod 10 or 15" behind the point of
rotation. It could traverse about 3" horizontally for each 20 degrees of
rotation and a photo would be taken. The overlap of the (say 30
degree) fields of any two consecutive shots is a stereo pair. So, a
moving camera capable of shooting burst or progressive mode could shoot
stereo in one single panoramic sweep. Many images would still be
required, maybe 20 or so. If a 50mm fl lens were used the curvature of
field would be minimal. Digital panoramic software tools allow image
morphing to correct for spherical distortion - but in stereo slight
image differences not due to parallax are a red flag. It is best not to
introduce a problem than to correct for it.
Traditional film panoramic cameras have an unusual perspective. One old
360 degree pan (Circuit camera photo)that I have shows milk wagons and
early dairy trucks circled around a park. All points on the horizon are
at the vanishing point.
The "node of emmission" of a lens is determined by rotating the lens on
a verticle axis ( when the lens is horizontal) there is a rotation point
where the image projected by the lens does not move. A slit was placed
behind the lens and traveled behind the lens as it rotated in front of
the curved film plane. Today the slit could be replaced by a scanner
sensor. The sensor "stares straight" all of the time. A pair could
stare straight and only aquire data on what is between the camera and
the vanishing point.
At least when images of this sort are made we will already have the
displays and software to view them. Film is wonderful stuff, and our
best images are certainly on film. But how does one view a stereo film
panorama without prismatic or reflective eyewear?
John Toeppen
http://members.home.net/toeppen
|