3
\$\begingroup\$

My ultimate goal is to set up some sort of digital strip camera rig for taking stereoscopic 360° panoramas.

I'd like to use a strip camera because it allows for fast panorama capture and theoretically by making more than one revolution you could make a stereo panoramic video, probably in time lapse (depending on how fast the strip camera is).

(General information on stereo pano photography is here: http://paulbourke.net/papers/vsmm2006/vsmm_talk.pdf)

However, the only options I've been able to find have been really expensive professional grade Photo Finish and panorama cameras, or phone camera apps (like ScanCamera on iOS) which are limited to 30 lines per second or less, meaning a low resolution (1024px wide) panorama would take 34 seconds to capture.

It's theoretically possible to capture a strip photography panorama using one 2D camera if it's set back a ways by taking a vertical strip of pixels from either side of the camera image.

After quite a bit of research I'm fairly certain there's no way to increase a phone camera's framerate enough to be usable for this. I don't have any hardware hacking experience, but I do have good high level programming skills.

Ideally such a system should ideally be able to capture at least 200 lines per second, and should cost less than $500. Film photography (such as the "Spinner 360" camera) is not a good option.

\$\endgroup\$
9
  • 2
    \$\begingroup\$ Sounds like you can use video from a 1080p capable camera and extract vertical lines from each frame. If you set up in portrait orientation you would get 1920 pixels per line and frame rates of up to 60 FPS on select models. \$\endgroup\$
    – Itai
    Commented Feb 7, 2013 at 16:27
  • 1
    \$\begingroup\$ BTW, some cameras already do this with larger strips. Both Sony and Fuji have Sweep and Motion panorama modes. On some, it will produce an 3D MPO file as well but for horizontal motion only. \$\endgroup\$
    – Itai
    Commented Feb 7, 2013 at 16:41
  • \$\begingroup\$ Hi Itai, an auto point and shoot panorama isn't what I'm looking for. Because of the particular way that stereoscopic panoramas are distorted, stitched panoramas don't work well without a ridiculous amount of photos. This is why I'm so particular about using a Strip Camera similar to "photo finish" cameras used for racing finish lines. \$\endgroup\$
    – Tim R.
    Commented Feb 7, 2013 at 17:09
  • \$\begingroup\$ Then what about the video option? You would be doing the stitching from extracted lines. Just making suggestions because I do not know if there is an answer to your question :) \$\endgroup\$
    – Itai
    Commented Feb 7, 2013 at 17:15
  • 1
    \$\begingroup\$ And by "you" I of course mean "I!" \$\endgroup\$
    – Tim R.
    Commented Feb 7, 2013 at 17:31

1 Answer 1

1
\$\begingroup\$

While you wish to capture the 360° panorama, you don't say what you want to do with the captured image. Do you wish to extract information from it or render the panorama for viewing?

There will be a finite time to make a sweep with a slit/strip aperture. The end of the captured record will not be seamless with the beginning of the sweep. There will be a mis-match unless the entire panorama can be captured simultaneously.

You can capture data with a panoramic view; but, the aspect ration will probably be unacceptable for aesthetic reasons.

Anamorphic lenses capture scenes with a distorted pictorial aspect ratio that is corrected by projecting the distorted image through the capture lens, backwards, to un-distort the image back to the proper proportions. A "matched set" of synchronized anamorphic cameras can be used to create the illusion of the effect you describe.

Another approach would be to use a reflecting (mirror) lens with the required distortion to capture the panorama and also be used to undo the distortion to project the image correctly to a viewer. The mirror used for this purpose is cone-shaped. By looking (or recording) directly along the axis at the apex of the cone, a panorama is revealed. By projecting the distorted scene upon the reflective cone, a panorama is projected perpendicularly to the axis of the reflected cone.

Here's an article to get you started with instructions to make what you need to prove your concept. Anamorphic experiments/theory - User's Guide

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.