Cool device for panoramic images

I haven't used one of these, but they look really cool. I just read about them on BBC's website. It would have been great to have one last fall when we were traveling out west:

Gigapan Systems Online - System Page GigaPan Systems

Tom

 

I have done stitched together photos for years, both static and VR movies. I usually shoot them by hand, but that's because I am cheap. I started with some commercial software, but now there are excellent open source versions available.

Tom

 

These are used for taking panoramic photos consisting of hundreds or thousands of individual photos. If you want to stand there and do it by hand, more power to you. Try not to get confused with the two dimensional overlaps.

Tom

 

You guys really don't have any imagination. If it were up to you, we would still be using wet plate photography.

The concept behind this sort of panorama is to have a wide field of view, but still be able to zoom in and see something a mile away. The technology was originally developed for use on Mars, where you want to be able to see the whole scene, but also have great detail. You can do it with very, very expensive imagers, or you can use a standard digital camera and shoot lots of photos. For use on Mars, weight is the main issue. For the rest of us, the cost of the camera is more important.

As an example, a panoramic photo using this technology was taken at the recent presidential inauguration. From this one photo, we can see the whole scene, but we can also zoom in and see individual faces.

To appreciate the concept, you have to think of interacting with a photo, not simply viewing it. You look at the whole, zoom in, pan around, zoom back out. Picture google earth, but with a single photo.

Another application is for very large format photos, but that is a more limited application.

Tom

 

Yes, true, but the specific application for this particular device was for the mars rover. When you have...um...shall we say the Hubble Telescope in orbit but pointed the other direction, you don't really need a robotic manipulator. The stitching software is still pretty important. Or perhaps you have a more down-to-earth application in mind? As for me, I was always more of a synthetic aperture type of guy.

Tom

 

Hugin batch stitcher does a good job with auto-gamma correction, as well as perspective distortions. Obviously you want to use a fixed exposure and fixed focus to minimize variation between frames. If you need more dynamic range, you can always combine multiple exposures at differing exposure settings. The resulting dynamic range can be staggering. Once again, this is another reason that an automated device is useful for these extreme images. Adding bracketed exposures multiplies the number of frames. You also want a really fast computer. Even with a speed demon, processing time is like sending out film to be developed.

Tom

 

Generally they are not taken at varying focal depths. Think of it as a mega-mega-mega pixel photo. You have a big photo, with extreme resolution, so you can zoom in to ridiculous levels.

Tom

 

You both are correct. There are two reasons you might want to use this approach, and the two reasons can be used separately or together:

1) High resolution - To take high resolution photos, you need a good lens and a high resolution imager (which could be digital or film). To increase the resolution of the imager, you need denser spacing of the detectors or a larger imager. A larger imager requires a bigger lens for the same effective f stop. Either way, it gets expensive to increase resolution, and there are limits to existing technology.

Photomosaic techniques break a photo into a number of smaller photos. Each individual photo is done at the full resolution of the camera. When these individual photos are stitched together (pasted edge to edge), the resulting photo is the combined resolution of the individual photos, minus any overlapping sections.

2 Perspective control - This is what Evan is describing, and it relates to wide angle and panoramic photos. When you look through a camera, you see a small piece of the world. This is the camera's field of view. By using a wide angle lens, you get a larger field of view at the expense of perspective distortion. Things start to look funny in the corners of your photo. Fish-eye lenses are an extreme example of this sort of distortion.

Doing a photomosaic allows the photographer to cover a wide field of view with a normal lens. Each photo covers a small field of view at a different angle. The photos are pasted together to make a large, wide angle or panoramic photo.

Google Earth is a good example of this sort of technology. Thousands upon thousands of photos are stitched together to cover the entire earth. You can sit back in space and view the entire earth, then zoom in and see your own house. Each photo was shot from overhead to keep the perspective consistent. The difference with Google Earth is that the camera was moved between shots, as opposed to a panorama where the camera is rotated between shots. Either way, the concept is the same. You can easily see that the same results could not be produced by taking one wide angle photo from a fixed point in space. There would be a lack of detail, you could only see one-half of the earth, and everything would be distorted as you neared the horizon.

Tom